{"success":true,"count":9,"items":[{"videoId":"xRh2sVcNXQ8","chunkIndex":0,"totalChunks":9,"title":"Marc Andreessen's 2026 Outlook: AI Timelines, US vs. China, and The Price of AI — Part 1 of 9","thumbnail":"https://i.ytimg.com/vi/xRh2sVcNXQ8/maxresdefault.jpg","duration":4878,"uploader":"a16z","youtubeUrl":"https://www.youtube.com/watch?v=xRh2sVcNXQ8","keywords":["artificial-intelligence","startup","venture-capital","technology-trends","business-models","silicon-valley","product-strategy","market-adoption","innovation","future-of-work"],"normalizedKeywords":["비즈니스·전략","프로덕트","기술 트렌드"],"targetAudience":[{"who":"초기 창업자","why":"AI 시장의 성장 속도와 비즈니스 모델 불확실성을 함께 이해할 수 있음"},{"who":"투자자","why":"AI의 상용화 속도와 투자 기회, 리스크를 판단하는 관점을 얻을 수 있음"},{"who":"프로덕트 매니저","why":"현재 AI 제품이 아직 초기 단계라는 인식과 제품 진화 방향을 배울 수 있음"}],"normalizedAudience":["창업자·스타트업","투자자·VC","프로덕트 매니저·기획자"],"summary":"이 영상은 AI 혁명을 인터넷보다 더 큰 기술 전환으로 규정하면서, 지금의 제품 형태가 5~10년 뒤에도 그대로일 가능성은 낮다고 본다. 마크 안드레센은 AI가 80년 넘게 이어진 인간 두뇌 모사 연구의 결실이며, 챗GPT 이후에야 비로소 대중화와 상용화가 동시에 폭발했다고 설명한다.\n\n동시에 그는 AI 사업의 핵심 쟁점이 '수익은 빠르게 늘지만 비용도 빠르게 늘어나는 구조'라고 짚으며, 소비자용·인프라용 두 가지 비즈니스 모델을 전제로 장기 경쟁이 전개될 것이라 말한다. 여론조사에서는 사람들이 AI를 불안해하지만 실제 행동은 정반대이며, 시장은 이미 빠르게 채택과 매출을 만들어내고 있다는 점을 강조한다.","insights":["AI는 인터넷보다 큰 범용 기술 전환으로 보아야 한다.","현재의 AI 제품 형태는 아직 최종 형태가 아니다.","검증된 뒤엔 경쟁자도 빠르게 따라붙는다.","여론의 공포보다 실제 사용 데이터가 더 중요하다.","AI 시장은 수익과 비용이 동시에 커지는 구조다."],"keyClips":[{"clipId":"xRh2sVcNXQ8:c0:15-21","startSegmentIndex":15,"endSegmentIndex":21,"startTime":102.799,"endTime":140.72,"durationSeconds":37.9,"preview":"AI는 역대급 전환","mustSee":true},{"clipId":"xRh2sVcNXQ8:c0:22-34","startSegmentIndex":22,"endSegmentIndex":34,"startTime":140.72,"endTime":283.199,"durationSeconds":142.5,"preview":"80년 숙성된 아이디어","mustSee":true},{"clipId":"xRh2sVcNXQ8:c0:35-41","startSegmentIndex":35,"endSegmentIndex":41,"startTime":283.199,"endTime":347.84000000000003,"durationSeconds":64.6,"preview":"실패 후 폭발","mustSee":false},{"clipId":"xRh2sVcNXQ8:c0:42-52","startSegmentIndex":42,"endSegmentIndex":52,"startTime":347.84000000000003,"endTime":462.319,"durationSeconds":114.5,"preview":"제품은 아직 초기","mustSee":false},{"clipId":"xRh2sVcNXQ8:c0:53-64","startSegmentIndex":53,"endSegmentIndex":64,"startTime":462.319,"endTime":552.32,"durationSeconds":90,"preview":"성장과 비용의 긴장","mustSee":false},{"clipId":"xRh2sVcNXQ8:c0:67-72","startSegmentIndex":67,"endSegmentIndex":72,"startTime":561.92,"endTime":597.68,"durationSeconds":35.8,"preview":"두 가지 사업모델","mustSee":false}],"curatedSegments":[{"segmentIndex":45,"text":"Um, venture We have our issues and venture but a huge advantage that we have is we don't have to we can bet on multiple strategies at the same time right um and we are doing this so we are betting on big models and small models and prepared train models and open source models right and you know and foundation models and applications right uh and consumer and enterprise and so the portfolio approach the nature of it is like we are aggressively basically uh we are aggressively investing behind every strategy that we've identified that we think has a plausible chance of even when that even when that's contradictory to another strategy that we're investing in and one is just like the world's messy and probably a bunch of things are going to work and so like there's not going to be clean yes or no answers to a bunch of this like a lot a answers to a bunch of this like a lot of the answers to this I think are just going to be and answers but the other is like if one of these strategies doesn't work like you know we're not trying to hedge per se but you know we're going to have representation in the portfolio of the alternate strategy and so we're going to have mult multiple ways to win.","startTime":3451.52,"endTime":3506.48,"durationSeconds":55,"level":"C1","overallScore":10,"rationale":"벤처 포트폴리오 전략의 핵심 논리 제시."},{"segmentIndex":34,"text":"Um whereas of course China exports just a tremendous number of physical things right um including like a huge part of like the entire supply chain of parts that basically go into everything that American manufacturers you know kind of make right and so by the time a US you know whatever by the time an American company brings a toy to market right or a uh you know or a car um or anything or a computer or a smartphone or whatever like it's got a lot of componentry in it that was made in China so there is a much tighter in interlinkage between the American and Chinese economies than there as the American and Soviet economies and you know may maybe you know Adam Smith or whatever might say you know that's good news for peace and that you know both countries need each other by the way the other part of that argument is that the Chinese basically the Chinese you know the Chinese governance model is based on high employment um you know because you know if you know at least all the geopolitical people say if China ended up with like 25 or 50% unemployment that would cause civil unrest which is the one thing that the CCP doesn't want and so the corresponding part of the trade pressure is China needs the American export market you know the American consumer is like a third of the global economy.","startTime":1535.52,"endTime":1595.84,"durationSeconds":60,"level":"C2","overallScore":8.8,"rationale":"무역 상호의존과 정치 논리를 깊게 설명."},{"segmentIndex":5,"text":"it came from a hedge fund and it like as far as we can tell it basically is this somewhat idiosyncratic situation where you just have this incredibly successful quant hedge fund with all these you know super geniuses um and the founder of that hedge fund you know basically decided to build AI um and you know at least external indications are this was a surprise to even the Chinese government it's impossible to prove you know what the Chinese government was surprised by or not but you know there's at least the atmospherics are that this was not exactly planned this was not a national champion tech company at the time that Deepseek was released it was it sort of came out of left field which by the way is very encouraging for the field that it was possible for somebody to do that kind of who was unknown right because it kind of means that maybe you don't need all these you know super genius superstar researchers maybe actually smart kids can just build this stuff which I think is the direction things are headed um and so that kicked off I would say like this kind of I don't know copycat's the wrong word but that was sort of it feels like the success of deepseek and the success of deepseek from China as open source kind of kicked off a sort of trend in China releasing these open source models um you know Look, the cynics, you know, in DC would say, you know, yeah, like they're dumping, right?","startTime":1839.12,"endTime":1900.72,"durationSeconds":62,"level":"C1","overallScore":9,"rationale":"배경 해석과 관점, 표현이 매우 풍부."},{"segmentIndex":60,"text":"um you know which is a key way that all this technology proliferates and so this law would have assigned downstream liability to any misuse of open source to the original developer of the open source and so you know you're an independent developer or you're an academic or you're a startup you develop and release an AI model the AI model works fine the day you release it it's great but like 5 years later it gets built into a nuclear power plant and then there's a meltdown of the nuclear","startTime":2379.76,"endTime":2389.76,"durationSeconds":10,"level":"C1","overallScore":9,"rationale":"오픈소스 책임 전가의 문제를 선명히 설명."},{"segmentIndex":27,"text":"Um, and so you've got like the most magic new technology in the world and then it's basically being served up by those companies in a as a cloud business and made basically available to everybody on the planet to just click and use and for like relatively small amounts of money and then on a usage basis which means and usage is great for startups because you it means you can start easily right you the you know there's very you know there's basically no fixed co for a startup building an AI app they don't have giant fixed cost because they could just tap into the open AI or anthropic or Google or Microsoft or whatever you know cloud you know tokens by the you know, intelligence tokens by the drink offering and just get going.","startTime":2607.44,"endTime":2639.839,"durationSeconds":32,"level":"C2","overallScore":9,"rationale":"스타트업 비용구조 통찰과 표현이 풍부."},{"segmentIndex":22,"text":"they're actually they actually as these products get more sophisticated they actually end up using many different kinds of models that are kind of customtailored to the specific aspects of how these products work. Um and so they may start out using one model but they end up using a dozen models and then in the fullness of time it might be 50 or 100 different models for different aspects of the product. Um and so they a lot of these the leading edge application companies are actually backward integrating and actually building their own AI models because they have the deepest understanding of their domain.","startTime":3222.48,"endTime":3234.72,"durationSeconds":12,"level":"C1","overallScore":9,"rationale":"앱 기업의 진화 논리를 깊게 설명."},{"segmentIndex":27,"text":"Um and so that you know that's I think >> small models though right Mark when you think about god models versus small models as you were describing that but that would be small would you categorize that as a small >> well some of them I mean we I will let them announce you know them I will let them announce you know whatever they're doing whenever it's appropriate but some of them are now also doing big model development um and again this is also part of what this is also part of the learning just in the last two years well so like here's a big learning just from the last two years which is very interesting which is two years ago or three years ago for sure you would have said wow open AI is like way out ahead um and like it's probably going to be impossible for anybody to catch up and then it's like okay well Anthropic caught up and so but you know they came out of open AI and so they had all the secrets you know whatever and so knew how to do it and so okay they caught up but surely nobody can catch up after them and then very quickly after that there were a raft of other companies that caught up very fast and XAI is maybe the best example of that which is like you know XAI you know Elon's company XAI is the company name gro is the consumer product version of it um XAI basically caught up to you know state-of-the-art openai anthropic level in like less than 12 months from a standing start right and So, and again that kind of argues against any kind of permanent lead, right, by any one incumbent that's just going to basically be able to lock the entire market down like if you can catch up like that.","startTime":3281.68,"endTime":3291.119,"durationSeconds":9,"level":"C1","overallScore":9,"rationale":"후발 추격 가능성에 대한 핵심 통찰."},{"segmentIndex":13,"text":"I you know my view is we need to be actually very respectful of that and we need to be very aware of that and basically that we you know I use the metaphor with the dog that caught the bus like we always wanted to work on things that matter we are working on things that matter uh people in the rest of society actually really do care about these things um and you know and it's our responsibility to think that all through very carefully and to do a good job um you know both not just building the technology but also explaining it you know look you know I think we have a real obligation to uh you know to really explain ourselves and engage on these issues um in terms of how to measure how going you know it's sort of the classic social science question um uh which is like okay if you want to understand basically you know patterns of people there's basically two ways to understand what people are doing and thinking um one is to ask them and then the other is to watch them um and like every social scientist like every sociologist will tell you this which basically is you can ask people right and the way you do that right is like you know surveys focus groups polls um you know what they think Um but then you can watch them and you then but then you can watch them and you can do what's you know called reveal preferences.","startTime":4316.88,"endTime":4379.28,"durationSeconds":62,"level":"C2","overallScore":9,"rationale":"책임론과 사회과학 틀 제시."},{"segmentIndex":52,"text":"Um and you know you have these like long elaborate you know discussions about you know theories on this and that and the other thing and then reality just like completely smacks you square in the face you know like you idiot right you know like you know what were you like you know this is like the you know the ultimate frustration of the business which is also very motivating which is the number of times that you think that you've applied superior analysis and then you've either invested or not invested based on that analysis and it turns out it was just you the analysis was just completely wrong right um and you know you just like completely overrated your ability to epistemically you know kind of analyze these things you just you know basically inflicted harm like I always the question is always you know it's sort of you know any activity that we do is it value add or is it actually value subtract right and I think in this business of all businesses is kind of like that and that applies to all of my own contributions as well so there is that and then I would say um you know maybe the final thing is just like I do have the entire internet ready to tell me that I'm an idiot so that also doesn't hurt and it does on a regular basis on the point of uh your alluding to earlier about uh decisions on investing in companies.","startTime":4683.36,"endTime":4748.96,"durationSeconds":66,"level":"C2","overallScore":9.2,"rationale":"현실 검증과 오판의 교훈이 강함."},{"segmentIndex":2,"text":"We're seeing companies grow much faster. I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in 5 or 10 years. I think things are going to get much more sophisticated from here.","startTime":8.72,"endTime":16.96,"durationSeconds":8,"level":"C1","overallScore":7.8,"rationale":"제품 진화에 대한 전망이 뚜렷함."},{"segmentIndex":4,"text":"These are trillion dollar questions, not answers. But once somebody proves that it's capable, it seems to not be that hard for other people to be able to catch up, even people with far less resources.","startTime":20.8,"endTime":29.84,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"기술 추격의 본질을 짚는 문장."},{"segmentIndex":9,"text":"One is to ask them and then the other is to watch them. And what you often see in many areas of human activity, including politics and many different aspects of society, the answers that you get when you ask people are very different than the answers that you get when you watch them.","startTime":52.64,"endTime":64.159,"durationSeconds":12,"level":"C1","overallScore":8,"rationale":"말과 행동의 괴리를 통찰함."},{"segmentIndex":27,"text":"Um and so like and that was of course the path that the industry took which was building these kind of hyper literal you know mathematical machines you know that could execute mathematical operations billions of times per second but of course had no ability to kind of deal with human beings the way humans like to be dealt with and so you know couldn't understand you know human speech human language um and so forth and that's the computer industry that got built over the last 80 years and that's the computer industry that's built all the wealth of uh and financial returns of the computer industry uh over the last 80 years you know across all the generations of computers from mainframes through to smartphones.","startTime":174.64,"endTime":206.8,"durationSeconds":32,"level":"C2","overallScore":8,"rationale":"컴퓨터 산업의 한계를 깊게 설명."},{"segmentIndex":34,"text":"Um but and the neural network basically didn't happen but the neural network as an idea continued to be explored in academia um and sort of advanced research by sort of a rump you know movement that was originally called cybernetics and then became known as artificial intelligence uh basically for the last 80 years and essentially it didn't work like essentially it was basically decade after decade of excessive optimism uh followed by disappointment.","startTime":260.639,"endTime":283.199,"durationSeconds":23,"level":"C2","overallScore":8,"rationale":"AI 역사 사이클을 압축해 설명."},{"segmentIndex":41,"text":"So, we're sort of three year we're sort of three years in um to, you know, basically what is effectively an 80-year revolution um of actually being able to deliver on all the promise that the people on the all the on the alternate path, the sort of human cognition model path, you know, kind of saw from the very beginning and then, you know, the great news with this technology is it's already it's kind of ultra democratized.","startTime":325.36,"endTime":347.84000000000003,"durationSeconds":22,"level":"C2","overallScore":8,"rationale":"AI 혁명의 위치와 성격을 압축."},{"segmentIndex":47,"text":"Like I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in five or 10 years.","startTime":423.199,"endTime":426.56,"durationSeconds":3,"level":"C1","overallScore":7.8,"rationale":"미래 제품 변화 전망이 선명함."},{"segmentIndex":15,"text":"Um, and so sort of the internet's the carrier wave for AI to be able to proliferate at kind of light speed uh into the broad base of the global population.","startTime":682.88,"endTime":691.6,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"AI 확산 비유가 강하고 표현도 인상적."},{"segmentIndex":16,"text":"And that's a let's just say that's a potential rate of proliferation of a new technology that's just far faster than has ever been possible before.","startTime":691.6,"endTime":697.8389999999999,"durationSeconds":6,"level":"B2","overallScore":7.4,"rationale":"기술 확산 속도에 대한 통찰이 강함."},{"segmentIndex":19,"text":"Um, you know, you couldn't download television, but you can download AI. Um, and then and an incredible rate. Um, and then they're monetizing really well. Um and the core business model, right, is actually quite interesting. Um and then as a consequence there's kind of this hyperdelation of per unit cost and then that is like driving you know just like you know a more than corresponding level of demand growth you know with elasticity.","startTime":704.8,"endTime":715.2,"durationSeconds":10,"level":"C1","overallScore":8,"rationale":"사업모델·수요 논리가 압축돼 통찰 높음."},{"segmentIndex":25,"text":"Um, and you know, if you have the ability to like inject more intelligence into your business and you have the ability to do, you know, even the most prosaic things like raise your customer service scores, uh, you know, increase upsells, um, uh, you know, or reduce churn or if you have the ability to, um, you know, run marketing campaigns more effectively, um, you know, all of which AI is directly relevant to, like, you know, these are like direct business payoffs, um, you know, that people are seeing already.","startTime":764.48,"endTime":786.639,"durationSeconds":22,"level":"C1","overallScore":8.2,"rationale":"기업 효용 사례와 실전 표현이 매우 풍부함."}],"generatedAt":"2026-06-24T23:18:18.812Z","keyClipsTotalSec":2283},{"videoId":"xRh2sVcNXQ8","chunkIndex":1,"totalChunks":9,"title":"Marc Andreessen's 2026 Outlook: AI Timelines, US vs. China, and The Price of AI — Part 2 of 9","thumbnail":"https://i.ytimg.com/vi/xRh2sVcNXQ8/maxresdefault.jpg","duration":4878,"uploader":"a16z","youtubeUrl":"https://www.youtube.com/watch?v=xRh2sVcNXQ8","keywords":["artificial-intelligence","machine-learning","ai-pricing","cloud-computing","semiconductors","enterprise-software","consumer-tech","open-source","market-dynamics"],"normalizedKeywords":["비즈니스·전략","기술 트렌드","프로덕트"],"targetAudience":[{"who":"창업자","why":"AI 제품의 가격 책정과 확장 구조를 이해하는 데 도움된다"},{"who":"투자자","why":"AI 인프라·모델 시장의 성장 논리와 비용 구조를 읽을 수 있다"},{"who":"프로덕트 매니저","why":"소비자·기업 고객에게 AI 가치가 어떻게 돈으로 바뀌는지 배울 수 있다"}],"normalizedAudience":["창업자·스타트업","투자자·VC","프로덕트 매니저·기획자"],"summary":"이 영상은 AI가 인터넷처럼 전 세계로 급속히 확산될 것이라는 낙관적 전망을 중심으로, 왜 AI의 가격이 빠르게 떨어지고 수요는 더 빠르게 늘어나는지 설명한다. 화자는 인터넷·스마트폰 보급의 역사와 비교하며, AI는 다운로드 가능한 기술이라 대중 확산 속도가 이전의 물리적 기술보다 훨씬 빠를 수 있다고 본다. 또한 소비자용 AI는 높은 구독 단가와 강한 사용성으로 수익화가 가능하고, 기업용 AI는 고객지원·업셀·이탈률 감소 같은 직접적 성과로 돈을 만든다고 주장한다.\n\n아울러 대형 모델과 소형 모델이 동시에 발전하면서, 고성능 모델은 계속 최상단에 남되 실제 배포량은 더 싸고 작은 모델이 넓게 퍼질 것이라고 전망한다. 중국의 오픈소스 모델 사례와 GPU의 수명 연장, 데이터센터·칩 공급의 부족이 결국 공급 확대로 이어지는 점도 언급하며, AI 산업 전반의 비용 하락과 투자 확대를 낙관한다.","insights":["다운로드 가능한 기술은 확산 속도가 물리 기술보다 압도적으로 빠르다.","AI의 가격 하락은 수요를 죽이기보다 오히려 폭발시킨다.","고성능 모델은 상단에 남고, 시장의 물량은 소형 모델이 가져간다.","기업이 사는 것은 모델이 아니라 더 높은 생산성과 매출이다.","칩·GPU 부족은 장기적으로 공급 확대와 비용 하락을 부른다."],"keyClips":[{"clipId":"xRh2sVcNXQ8:c1:1-19","startSegmentIndex":1,"endSegmentIndex":19,"startTime":601.59,"endTime":715.2,"durationSeconds":113.6,"preview":"인터넷급 확산 논리","mustSee":true},{"clipId":"xRh2sVcNXQ8:c1:20-23","startSegmentIndex":20,"endSegmentIndex":23,"startTime":715.2,"endTime":758.24,"durationSeconds":43,"preview":"소비자 과금의 확장","mustSee":false},{"clipId":"xRh2sVcNXQ8:c1:24-36","startSegmentIndex":24,"endSegmentIndex":36,"startTime":758.24,"endTime":941.12,"durationSeconds":182.9,"preview":"기업용 AI의 본질","mustSee":true},{"clipId":"xRh2sVcNXQ8:c1:39-52","startSegmentIndex":39,"endSegmentIndex":52,"startTime":952.24,"endTime":1109.12,"durationSeconds":156.9,"preview":"작은 모델의 반격","mustSee":true},{"clipId":"xRh2sVcNXQ8:c1:54-61","startSegmentIndex":54,"endSegmentIndex":61,"startTime":1125.12,"endTime":1200.799,"durationSeconds":75.7,"preview":"AI 산업의 구조","mustSee":false}],"curatedSegments":[{"segmentIndex":45,"text":"Um, venture We have our issues and venture but a huge advantage that we have is we don't have to we can bet on multiple strategies at the same time right um and we are doing this so we are betting on big models and small models and prepared train models and open source models right and you know and foundation models and applications right uh and consumer and enterprise and so the portfolio approach the nature of it is like we are aggressively basically uh we are aggressively investing behind every strategy that we've identified that we think has a plausible chance of even when that even when that's contradictory to another strategy that we're investing in and one is just like the world's messy and probably a bunch of things are going to work and so like there's not going to be clean yes or no answers to a bunch of this like a lot a answers to a bunch of this like a lot of the answers to this I think are just going to be and answers but the other is like if one of these strategies doesn't work like you know we're not trying to hedge per se but you know we're going to have representation in the portfolio of the alternate strategy and so we're going to have mult multiple ways to win.","startTime":3451.52,"endTime":3506.48,"durationSeconds":55,"level":"C1","overallScore":10,"rationale":"벤처 포트폴리오 전략의 핵심 논리 제시."},{"segmentIndex":34,"text":"Um whereas of course China exports just a tremendous number of physical things right um including like a huge part of like the entire supply chain of parts that basically go into everything that American manufacturers you know kind of make right and so by the time a US you know whatever by the time an American company brings a toy to market right or a uh you know or a car um or anything or a computer or a smartphone or whatever like it's got a lot of componentry in it that was made in China so there is a much tighter in interlinkage between the American and Chinese economies than there as the American and Soviet economies and you know may maybe you know Adam Smith or whatever might say you know that's good news for peace and that you know both countries need each other by the way the other part of that argument is that the Chinese basically the Chinese you know the Chinese governance model is based on high employment um you know because you know if you know at least all the geopolitical people say if China ended up with like 25 or 50% unemployment that would cause civil unrest which is the one thing that the CCP doesn't want and so the corresponding part of the trade pressure is China needs the American export market you know the American consumer is like a third of the global economy.","startTime":1535.52,"endTime":1595.84,"durationSeconds":60,"level":"C2","overallScore":8.8,"rationale":"무역 상호의존과 정치 논리를 깊게 설명."},{"segmentIndex":5,"text":"it came from a hedge fund and it like as far as we can tell it basically is this somewhat idiosyncratic situation where you just have this incredibly successful quant hedge fund with all these you know super geniuses um and the founder of that hedge fund you know basically decided to build AI um and you know at least external indications are this was a surprise to even the Chinese government it's impossible to prove you know what the Chinese government was surprised by or not but you know there's at least the atmospherics are that this was not exactly planned this was not a national champion tech company at the time that Deepseek was released it was it sort of came out of left field which by the way is very encouraging for the field that it was possible for somebody to do that kind of who was unknown right because it kind of means that maybe you don't need all these you know super genius superstar researchers maybe actually smart kids can just build this stuff which I think is the direction things are headed um and so that kicked off I would say like this kind of I don't know copycat's the wrong word but that was sort of it feels like the success of deepseek and the success of deepseek from China as open source kind of kicked off a sort of trend in China releasing these open source models um you know Look, the cynics, you know, in DC would say, you know, yeah, like they're dumping, right?","startTime":1839.12,"endTime":1900.72,"durationSeconds":62,"level":"C1","overallScore":9,"rationale":"배경 해석과 관점, 표현이 매우 풍부."},{"segmentIndex":60,"text":"um you know which is a key way that all this technology proliferates and so this law would have assigned downstream liability to any misuse of open source to the original developer of the open source and so you know you're an independent developer or you're an academic or you're a startup you develop and release an AI model the AI model works fine the day you release it it's great but like 5 years later it gets built into a nuclear power plant and then there's a meltdown of the nuclear","startTime":2379.76,"endTime":2389.76,"durationSeconds":10,"level":"C1","overallScore":9,"rationale":"오픈소스 책임 전가의 문제를 선명히 설명."},{"segmentIndex":27,"text":"Um, and so you've got like the most magic new technology in the world and then it's basically being served up by those companies in a as a cloud business and made basically available to everybody on the planet to just click and use and for like relatively small amounts of money and then on a usage basis which means and usage is great for startups because you it means you can start easily right you the you know there's very you know there's basically no fixed co for a startup building an AI app they don't have giant fixed cost because they could just tap into the open AI or anthropic or Google or Microsoft or whatever you know cloud you know tokens by the you know, intelligence tokens by the drink offering and just get going.","startTime":2607.44,"endTime":2639.839,"durationSeconds":32,"level":"C2","overallScore":9,"rationale":"스타트업 비용구조 통찰과 표현이 풍부."},{"segmentIndex":22,"text":"they're actually they actually as these products get more sophisticated they actually end up using many different kinds of models that are kind of customtailored to the specific aspects of how these products work. Um and so they may start out using one model but they end up using a dozen models and then in the fullness of time it might be 50 or 100 different models for different aspects of the product. Um and so they a lot of these the leading edge application companies are actually backward integrating and actually building their own AI models because they have the deepest understanding of their domain.","startTime":3222.48,"endTime":3234.72,"durationSeconds":12,"level":"C1","overallScore":9,"rationale":"앱 기업의 진화 논리를 깊게 설명."},{"segmentIndex":27,"text":"Um and so that you know that's I think >> small models though right Mark when you think about god models versus small models as you were describing that but that would be small would you categorize that as a small >> well some of them I mean we I will let them announce you know them I will let them announce you know whatever they're doing whenever it's appropriate but some of them are now also doing big model development um and again this is also part of what this is also part of the learning just in the last two years well so like here's a big learning just from the last two years which is very interesting which is two years ago or three years ago for sure you would have said wow open AI is like way out ahead um and like it's probably going to be impossible for anybody to catch up and then it's like okay well Anthropic caught up and so but you know they came out of open AI and so they had all the secrets you know whatever and so knew how to do it and so okay they caught up but surely nobody can catch up after them and then very quickly after that there were a raft of other companies that caught up very fast and XAI is maybe the best example of that which is like you know XAI you know Elon's company XAI is the company name gro is the consumer product version of it um XAI basically caught up to you know state-of-the-art openai anthropic level in like less than 12 months from a standing start right and So, and again that kind of argues against any kind of permanent lead, right, by any one incumbent that's just going to basically be able to lock the entire market down like if you can catch up like that.","startTime":3281.68,"endTime":3291.119,"durationSeconds":9,"level":"C1","overallScore":9,"rationale":"후발 추격 가능성에 대한 핵심 통찰."},{"segmentIndex":13,"text":"I you know my view is we need to be actually very respectful of that and we need to be very aware of that and basically that we you know I use the metaphor with the dog that caught the bus like we always wanted to work on things that matter we are working on things that matter uh people in the rest of society actually really do care about these things um and you know and it's our responsibility to think that all through very carefully and to do a good job um you know both not just building the technology but also explaining it you know look you know I think we have a real obligation to uh you know to really explain ourselves and engage on these issues um in terms of how to measure how going you know it's sort of the classic social science question um uh which is like okay if you want to understand basically you know patterns of people there's basically two ways to understand what people are doing and thinking um one is to ask them and then the other is to watch them um and like every social scientist like every sociologist will tell you this which basically is you can ask people right and the way you do that right is like you know surveys focus groups polls um you know what they think Um but then you can watch them and you then but then you can watch them and you can do what's you know called reveal preferences.","startTime":4316.88,"endTime":4379.28,"durationSeconds":62,"level":"C2","overallScore":9,"rationale":"책임론과 사회과학 틀 제시."},{"segmentIndex":52,"text":"Um and you know you have these like long elaborate you know discussions about you know theories on this and that and the other thing and then reality just like completely smacks you square in the face you know like you idiot right you know like you know what were you like you know this is like the you know the ultimate frustration of the business which is also very motivating which is the number of times that you think that you've applied superior analysis and then you've either invested or not invested based on that analysis and it turns out it was just you the analysis was just completely wrong right um and you know you just like completely overrated your ability to epistemically you know kind of analyze these things you just you know basically inflicted harm like I always the question is always you know it's sort of you know any activity that we do is it value add or is it actually value subtract right and I think in this business of all businesses is kind of like that and that applies to all of my own contributions as well so there is that and then I would say um you know maybe the final thing is just like I do have the entire internet ready to tell me that I'm an idiot so that also doesn't hurt and it does on a regular basis on the point of uh your alluding to earlier about uh decisions on investing in companies.","startTime":4683.36,"endTime":4748.96,"durationSeconds":66,"level":"C2","overallScore":9.2,"rationale":"현실 검증과 오판의 교훈이 강함."},{"segmentIndex":2,"text":"We're seeing companies grow much faster. I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in 5 or 10 years. I think things are going to get much more sophisticated from here.","startTime":8.72,"endTime":16.96,"durationSeconds":8,"level":"C1","overallScore":7.8,"rationale":"제품 진화에 대한 전망이 뚜렷함."},{"segmentIndex":4,"text":"These are trillion dollar questions, not answers. But once somebody proves that it's capable, it seems to not be that hard for other people to be able to catch up, even people with far less resources.","startTime":20.8,"endTime":29.84,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"기술 추격의 본질을 짚는 문장."},{"segmentIndex":9,"text":"One is to ask them and then the other is to watch them. And what you often see in many areas of human activity, including politics and many different aspects of society, the answers that you get when you ask people are very different than the answers that you get when you watch them.","startTime":52.64,"endTime":64.159,"durationSeconds":12,"level":"C1","overallScore":8,"rationale":"말과 행동의 괴리를 통찰함."},{"segmentIndex":27,"text":"Um and so like and that was of course the path that the industry took which was building these kind of hyper literal you know mathematical machines you know that could execute mathematical operations billions of times per second but of course had no ability to kind of deal with human beings the way humans like to be dealt with and so you know couldn't understand you know human speech human language um and so forth and that's the computer industry that got built over the last 80 years and that's the computer industry that's built all the wealth of uh and financial returns of the computer industry uh over the last 80 years you know across all the generations of computers from mainframes through to smartphones.","startTime":174.64,"endTime":206.8,"durationSeconds":32,"level":"C2","overallScore":8,"rationale":"컴퓨터 산업의 한계를 깊게 설명."},{"segmentIndex":34,"text":"Um but and the neural network basically didn't happen but the neural network as an idea continued to be explored in academia um and sort of advanced research by sort of a rump you know movement that was originally called cybernetics and then became known as artificial intelligence uh basically for the last 80 years and essentially it didn't work like essentially it was basically decade after decade of excessive optimism uh followed by disappointment.","startTime":260.639,"endTime":283.199,"durationSeconds":23,"level":"C2","overallScore":8,"rationale":"AI 역사 사이클을 압축해 설명."},{"segmentIndex":41,"text":"So, we're sort of three year we're sort of three years in um to, you know, basically what is effectively an 80-year revolution um of actually being able to deliver on all the promise that the people on the all the on the alternate path, the sort of human cognition model path, you know, kind of saw from the very beginning and then, you know, the great news with this technology is it's already it's kind of ultra democratized.","startTime":325.36,"endTime":347.84000000000003,"durationSeconds":22,"level":"C2","overallScore":8,"rationale":"AI 혁명의 위치와 성격을 압축."},{"segmentIndex":47,"text":"Like I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in five or 10 years.","startTime":423.199,"endTime":426.56,"durationSeconds":3,"level":"C1","overallScore":7.8,"rationale":"미래 제품 변화 전망이 선명함."},{"segmentIndex":15,"text":"Um, and so sort of the internet's the carrier wave for AI to be able to proliferate at kind of light speed uh into the broad base of the global population.","startTime":682.88,"endTime":691.6,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"AI 확산 비유가 강하고 표현도 인상적."},{"segmentIndex":16,"text":"And that's a let's just say that's a potential rate of proliferation of a new technology that's just far faster than has ever been possible before.","startTime":691.6,"endTime":697.8389999999999,"durationSeconds":6,"level":"B2","overallScore":7.4,"rationale":"기술 확산 속도에 대한 통찰이 강함."},{"segmentIndex":19,"text":"Um, you know, you couldn't download television, but you can download AI. Um, and then and an incredible rate. Um, and then they're monetizing really well. Um and the core business model, right, is actually quite interesting. Um and then as a consequence there's kind of this hyperdelation of per unit cost and then that is like driving you know just like you know a more than corresponding level of demand growth you know with elasticity.","startTime":704.8,"endTime":715.2,"durationSeconds":10,"level":"C1","overallScore":8,"rationale":"사업모델·수요 논리가 압축돼 통찰 높음."},{"segmentIndex":25,"text":"Um, and you know, if you have the ability to like inject more intelligence into your business and you have the ability to do, you know, even the most prosaic things like raise your customer service scores, uh, you know, increase upsells, um, uh, you know, or reduce churn or if you have the ability to, um, you know, run marketing campaigns more effectively, um, you know, all of which AI is directly relevant to, like, you know, these are like direct business payoffs, um, you know, that people are seeing already.","startTime":764.48,"endTime":786.639,"durationSeconds":22,"level":"C1","overallScore":8.2,"rationale":"기업 효용 사례와 실전 표현이 매우 풍부함."}],"generatedAt":"2026-06-24T23:18:51.870Z","keyClipsTotalSec":2283},{"videoId":"xRh2sVcNXQ8","chunkIndex":2,"totalChunks":9,"title":"Marc Andreessen's 2026 Outlook: AI Timelines, US vs. China, and The Price of AI — Part 3 of 9","thumbnail":"https://i.ytimg.com/vi/xRh2sVcNXQ8/maxresdefault.jpg","duration":4878,"uploader":"a16z","youtubeUrl":"https://www.youtube.com/watch?v=xRh2sVcNXQ8","keywords":["ai","semiconductors","chips","nvidia","china","geopolitics","robotics","hardware","open-source-ai","technology-trends"],"normalizedKeywords":["기술 트렌드","비즈니스·전략","엔지니어링"],"targetAudience":[{"who":"투자자","why":"AI 칩 가격, 공급 경쟁, 산업 구조 변화가 투자 기회로 이어짐"},{"who":"창업자","why":"AI·칩·로보틱스의 다음 수요 지형을 읽는 데 도움이 됨"},{"who":"엔지니어","why":"GPU와 전용 AI 칩의 구조적 차이를 이해하는 데 유용함"}],"normalizedAudience":["투자자·VC","창업자·스타트업","엔지니어·개발자"],"summary":"이 영상은 AI 칩 시장이 어떻게 독점에서 경쟁과 상품화 국면으로 이동하는지, 그리고 그 변화가 Nvidia와 AI 산업 전반에 어떤 의미를 갖는지를 설명한다. 화자는 GPU가 AI의 핵심 인프라가 된 배경을 PC 시대의 CPU/GPU 분업 구조에서 풀어내며, 오늘날의 AI 칩 경쟁이 사실상 전용 AI 칩, 하이퍼스케일러의 자체 칩, 중국의 추격이 얽힌 다자 경쟁이라고 본다.\n\n또한 이 논의는 단순한 기술 이야기를 넘어 미·중 관계, 공급망, 로보틱스 경쟁까지 이어진다. 미국과 중국은 서로 깊게 연결된 경제 구조를 갖고 있지만, 동시에 AI가 세계로 퍼질 때 어느 쪽의 표준이 지배할지를 두고 사실상의 전략 경쟁을 벌이고 있다고 정리한다. 전반적으로 이 영상은 AI의 '모델'보다 '칩과 공급망'이 앞으로 더 큰 변수가 될 수 있다는 관점을 제시한다.","insights":["기술 독점의 종착점은 경쟁자 유입과 가격 하락이다.","AI 칩의 가치는 모델보다 공급망과 제조능력에서 갈린다.","범용 GPU의 시대 뒤에는 전용 AI 칩 경쟁이 온다.","미중 AI 경쟁은 기술전이자 글로벌 표준 싸움이다.","로보틱스는 AI보다 더 공급망 중심의 경쟁이 된다."],"keyClips":[{"clipId":"xRh2sVcNXQ8:c2:1-7","startSegmentIndex":1,"endSegmentIndex":7,"startTime":1200.95,"endTime":1266.88,"durationSeconds":65.9,"preview":"칩 독점의 끝","mustSee":false},{"clipId":"xRh2sVcNXQ8:c2:8-19","startSegmentIndex":8,"endSegmentIndex":19,"startTime":1266.88,"endTime":1428.08,"durationSeconds":161.2,"preview":"GPU에서 전용칩","mustSee":true},{"clipId":"xRh2sVcNXQ8:c2:20-25","startSegmentIndex":20,"endSegmentIndex":25,"startTime":1428.08,"endTime":1456.08,"durationSeconds":28,"preview":"중국 칩 생태계","mustSee":false},{"clipId":"xRh2sVcNXQ8:c2:26-41","startSegmentIndex":26,"endSegmentIndex":41,"startTime":1456.08,"endTime":1706.96,"durationSeconds":250.9,"preview":"미중 AI 냉전","mustSee":true},{"clipId":"xRh2sVcNXQ8:c2:42-50","startSegmentIndex":42,"endSegmentIndex":50,"startTime":1706.96,"endTime":1807.44,"durationSeconds":100.5,"preview":"로봇 경쟁의 시작","mustSee":false}],"curatedSegments":[{"segmentIndex":45,"text":"Um, venture We have our issues and venture but a huge advantage that we have is we don't have to we can bet on multiple strategies at the same time right um and we are doing this so we are betting on big models and small models and prepared train models and open source models right and you know and foundation models and applications right uh and consumer and enterprise and so the portfolio approach the nature of it is like we are aggressively basically uh we are aggressively investing behind every strategy that we've identified that we think has a plausible chance of even when that even when that's contradictory to another strategy that we're investing in and one is just like the world's messy and probably a bunch of things are going to work and so like there's not going to be clean yes or no answers to a bunch of this like a lot a answers to a bunch of this like a lot of the answers to this I think are just going to be and answers but the other is like if one of these strategies doesn't work like you know we're not trying to hedge per se but you know we're going to have representation in the portfolio of the alternate strategy and so we're going to have mult multiple ways to win.","startTime":3451.52,"endTime":3506.48,"durationSeconds":55,"level":"C1","overallScore":10,"rationale":"벤처 포트폴리오 전략의 핵심 논리 제시."},{"segmentIndex":34,"text":"Um whereas of course China exports just a tremendous number of physical things right um including like a huge part of like the entire supply chain of parts that basically go into everything that American manufacturers you know kind of make right and so by the time a US you know whatever by the time an American company brings a toy to market right or a uh you know or a car um or anything or a computer or a smartphone or whatever like it's got a lot of componentry in it that was made in China so there is a much tighter in interlinkage between the American and Chinese economies than there as the American and Soviet economies and you know may maybe you know Adam Smith or whatever might say you know that's good news for peace and that you know both countries need each other by the way the other part of that argument is that the Chinese basically the Chinese you know the Chinese governance model is based on high employment um you know because you know if you know at least all the geopolitical people say if China ended up with like 25 or 50% unemployment that would cause civil unrest which is the one thing that the CCP doesn't want and so the corresponding part of the trade pressure is China needs the American export market you know the American consumer is like a third of the global economy.","startTime":1535.52,"endTime":1595.84,"durationSeconds":60,"level":"C2","overallScore":8.8,"rationale":"무역 상호의존과 정치 논리를 깊게 설명."},{"segmentIndex":5,"text":"it came from a hedge fund and it like as far as we can tell it basically is this somewhat idiosyncratic situation where you just have this incredibly successful quant hedge fund with all these you know super geniuses um and the founder of that hedge fund you know basically decided to build AI um and you know at least external indications are this was a surprise to even the Chinese government it's impossible to prove you know what the Chinese government was surprised by or not but you know there's at least the atmospherics are that this was not exactly planned this was not a national champion tech company at the time that Deepseek was released it was it sort of came out of left field which by the way is very encouraging for the field that it was possible for somebody to do that kind of who was unknown right because it kind of means that maybe you don't need all these you know super genius superstar researchers maybe actually smart kids can just build this stuff which I think is the direction things are headed um and so that kicked off I would say like this kind of I don't know copycat's the wrong word but that was sort of it feels like the success of deepseek and the success of deepseek from China as open source kind of kicked off a sort of trend in China releasing these open source models um you know Look, the cynics, you know, in DC would say, you know, yeah, like they're dumping, right?","startTime":1839.12,"endTime":1900.72,"durationSeconds":62,"level":"C1","overallScore":9,"rationale":"배경 해석과 관점, 표현이 매우 풍부."},{"segmentIndex":60,"text":"um you know which is a key way that all this technology proliferates and so this law would have assigned downstream liability to any misuse of open source to the original developer of the open source and so you know you're an independent developer or you're an academic or you're a startup you develop and release an AI model the AI model works fine the day you release it it's great but like 5 years later it gets built into a nuclear power plant and then there's a meltdown of the nuclear","startTime":2379.76,"endTime":2389.76,"durationSeconds":10,"level":"C1","overallScore":9,"rationale":"오픈소스 책임 전가의 문제를 선명히 설명."},{"segmentIndex":27,"text":"Um, and so you've got like the most magic new technology in the world and then it's basically being served up by those companies in a as a cloud business and made basically available to everybody on the planet to just click and use and for like relatively small amounts of money and then on a usage basis which means and usage is great for startups because you it means you can start easily right you the you know there's very you know there's basically no fixed co for a startup building an AI app they don't have giant fixed cost because they could just tap into the open AI or anthropic or Google or Microsoft or whatever you know cloud you know tokens by the you know, intelligence tokens by the drink offering and just get going.","startTime":2607.44,"endTime":2639.839,"durationSeconds":32,"level":"C2","overallScore":9,"rationale":"스타트업 비용구조 통찰과 표현이 풍부."},{"segmentIndex":22,"text":"they're actually they actually as these products get more sophisticated they actually end up using many different kinds of models that are kind of customtailored to the specific aspects of how these products work. Um and so they may start out using one model but they end up using a dozen models and then in the fullness of time it might be 50 or 100 different models for different aspects of the product. Um and so they a lot of these the leading edge application companies are actually backward integrating and actually building their own AI models because they have the deepest understanding of their domain.","startTime":3222.48,"endTime":3234.72,"durationSeconds":12,"level":"C1","overallScore":9,"rationale":"앱 기업의 진화 논리를 깊게 설명."},{"segmentIndex":27,"text":"Um and so that you know that's I think >> small models though right Mark when you think about god models versus small models as you were describing that but that would be small would you categorize that as a small >> well some of them I mean we I will let them announce you know them I will let them announce you know whatever they're doing whenever it's appropriate but some of them are now also doing big model development um and again this is also part of what this is also part of the learning just in the last two years well so like here's a big learning just from the last two years which is very interesting which is two years ago or three years ago for sure you would have said wow open AI is like way out ahead um and like it's probably going to be impossible for anybody to catch up and then it's like okay well Anthropic caught up and so but you know they came out of open AI and so they had all the secrets you know whatever and so knew how to do it and so okay they caught up but surely nobody can catch up after them and then very quickly after that there were a raft of other companies that caught up very fast and XAI is maybe the best example of that which is like you know XAI you know Elon's company XAI is the company name gro is the consumer product version of it um XAI basically caught up to you know state-of-the-art openai anthropic level in like less than 12 months from a standing start right and So, and again that kind of argues against any kind of permanent lead, right, by any one incumbent that's just going to basically be able to lock the entire market down like if you can catch up like that.","startTime":3281.68,"endTime":3291.119,"durationSeconds":9,"level":"C1","overallScore":9,"rationale":"후발 추격 가능성에 대한 핵심 통찰."},{"segmentIndex":13,"text":"I you know my view is we need to be actually very respectful of that and we need to be very aware of that and basically that we you know I use the metaphor with the dog that caught the bus like we always wanted to work on things that matter we are working on things that matter uh people in the rest of society actually really do care about these things um and you know and it's our responsibility to think that all through very carefully and to do a good job um you know both not just building the technology but also explaining it you know look you know I think we have a real obligation to uh you know to really explain ourselves and engage on these issues um in terms of how to measure how going you know it's sort of the classic social science question um uh which is like okay if you want to understand basically you know patterns of people there's basically two ways to understand what people are doing and thinking um one is to ask them and then the other is to watch them um and like every social scientist like every sociologist will tell you this which basically is you can ask people right and the way you do that right is like you know surveys focus groups polls um you know what they think Um but then you can watch them and you then but then you can watch them and you can do what's you know called reveal preferences.","startTime":4316.88,"endTime":4379.28,"durationSeconds":62,"level":"C2","overallScore":9,"rationale":"책임론과 사회과학 틀 제시."},{"segmentIndex":52,"text":"Um and you know you have these like long elaborate you know discussions about you know theories on this and that and the other thing and then reality just like completely smacks you square in the face you know like you idiot right you know like you know what were you like you know this is like the you know the ultimate frustration of the business which is also very motivating which is the number of times that you think that you've applied superior analysis and then you've either invested or not invested based on that analysis and it turns out it was just you the analysis was just completely wrong right um and you know you just like completely overrated your ability to epistemically you know kind of analyze these things you just you know basically inflicted harm like I always the question is always you know it's sort of you know any activity that we do is it value add or is it actually value subtract right and I think in this business of all businesses is kind of like that and that applies to all of my own contributions as well so there is that and then I would say um you know maybe the final thing is just like I do have the entire internet ready to tell me that I'm an idiot so that also doesn't hurt and it does on a regular basis on the point of uh your alluding to earlier about uh decisions on investing in companies.","startTime":4683.36,"endTime":4748.96,"durationSeconds":66,"level":"C2","overallScore":9.2,"rationale":"현실 검증과 오판의 교훈이 강함."},{"segmentIndex":2,"text":"We're seeing companies grow much faster. I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in 5 or 10 years. I think things are going to get much more sophisticated from here.","startTime":8.72,"endTime":16.96,"durationSeconds":8,"level":"C1","overallScore":7.8,"rationale":"제품 진화에 대한 전망이 뚜렷함."},{"segmentIndex":4,"text":"These are trillion dollar questions, not answers. But once somebody proves that it's capable, it seems to not be that hard for other people to be able to catch up, even people with far less resources.","startTime":20.8,"endTime":29.84,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"기술 추격의 본질을 짚는 문장."},{"segmentIndex":9,"text":"One is to ask them and then the other is to watch them. And what you often see in many areas of human activity, including politics and many different aspects of society, the answers that you get when you ask people are very different than the answers that you get when you watch them.","startTime":52.64,"endTime":64.159,"durationSeconds":12,"level":"C1","overallScore":8,"rationale":"말과 행동의 괴리를 통찰함."},{"segmentIndex":27,"text":"Um and so like and that was of course the path that the industry took which was building these kind of hyper literal you know mathematical machines you know that could execute mathematical operations billions of times per second but of course had no ability to kind of deal with human beings the way humans like to be dealt with and so you know couldn't understand you know human speech human language um and so forth and that's the computer industry that got built over the last 80 years and that's the computer industry that's built all the wealth of uh and financial returns of the computer industry uh over the last 80 years you know across all the generations of computers from mainframes through to smartphones.","startTime":174.64,"endTime":206.8,"durationSeconds":32,"level":"C2","overallScore":8,"rationale":"컴퓨터 산업의 한계를 깊게 설명."},{"segmentIndex":34,"text":"Um but and the neural network basically didn't happen but the neural network as an idea continued to be explored in academia um and sort of advanced research by sort of a rump you know movement that was originally called cybernetics and then became known as artificial intelligence uh basically for the last 80 years and essentially it didn't work like essentially it was basically decade after decade of excessive optimism uh followed by disappointment.","startTime":260.639,"endTime":283.199,"durationSeconds":23,"level":"C2","overallScore":8,"rationale":"AI 역사 사이클을 압축해 설명."},{"segmentIndex":41,"text":"So, we're sort of three year we're sort of three years in um to, you know, basically what is effectively an 80-year revolution um of actually being able to deliver on all the promise that the people on the all the on the alternate path, the sort of human cognition model path, you know, kind of saw from the very beginning and then, you know, the great news with this technology is it's already it's kind of ultra democratized.","startTime":325.36,"endTime":347.84000000000003,"durationSeconds":22,"level":"C2","overallScore":8,"rationale":"AI 혁명의 위치와 성격을 압축."},{"segmentIndex":47,"text":"Like I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in five or 10 years.","startTime":423.199,"endTime":426.56,"durationSeconds":3,"level":"C1","overallScore":7.8,"rationale":"미래 제품 변화 전망이 선명함."},{"segmentIndex":15,"text":"Um, and so sort of the internet's the carrier wave for AI to be able to proliferate at kind of light speed uh into the broad base of the global population.","startTime":682.88,"endTime":691.6,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"AI 확산 비유가 강하고 표현도 인상적."},{"segmentIndex":16,"text":"And that's a let's just say that's a potential rate of proliferation of a new technology that's just far faster than has ever been possible before.","startTime":691.6,"endTime":697.8389999999999,"durationSeconds":6,"level":"B2","overallScore":7.4,"rationale":"기술 확산 속도에 대한 통찰이 강함."},{"segmentIndex":19,"text":"Um, you know, you couldn't download television, but you can download AI. Um, and then and an incredible rate. Um, and then they're monetizing really well. Um and the core business model, right, is actually quite interesting. Um and then as a consequence there's kind of this hyperdelation of per unit cost and then that is like driving you know just like you know a more than corresponding level of demand growth you know with elasticity.","startTime":704.8,"endTime":715.2,"durationSeconds":10,"level":"C1","overallScore":8,"rationale":"사업모델·수요 논리가 압축돼 통찰 높음."},{"segmentIndex":25,"text":"Um, and you know, if you have the ability to like inject more intelligence into your business and you have the ability to do, you know, even the most prosaic things like raise your customer service scores, uh, you know, increase upsells, um, uh, you know, or reduce churn or if you have the ability to, um, you know, run marketing campaigns more effectively, um, you know, all of which AI is directly relevant to, like, you know, these are like direct business payoffs, um, you know, that people are seeing already.","startTime":764.48,"endTime":786.639,"durationSeconds":22,"level":"C1","overallScore":8.2,"rationale":"기업 효용 사례와 실전 표현이 매우 풍부함."}],"generatedAt":"2026-06-24T23:19:26.667Z","keyClipsTotalSec":2283},{"videoId":"xRh2sVcNXQ8","chunkIndex":3,"totalChunks":9,"title":"Marc Andreessen's 2026 Outlook: AI Timelines, US vs. China, and The Price of AI — Part 4 of 9","thumbnail":"https://i.ytimg.com/vi/xRh2sVcNXQ8/maxresdefault.jpg","duration":4878,"uploader":"a16z","youtubeUrl":"https://www.youtube.com/watch?v=xRh2sVcNXQ8","keywords":["ai-policy","regulation","china","europe","open-source","artificial-intelligence","industrial-policy","geopolitics","startup"],"normalizedKeywords":["비즈니스·전략","기술 트렌드","리더십·매니지먼트"],"targetAudience":[{"who":"창업자","why":"AI 규제와 오픈소스 책임 문제를 이해해야 제품 전략을 세울 수 있음"},{"who":"투자자","why":"미·중 AI 경쟁과 정책 리스크가 시장 구조를 바꾸는 방식을 볼 수 있음"},{"who":"정책 관심자","why":"연방과 주 정부, EU 규제가 AI 산업에 미치는 영향을 파악할 수 있음"}],"normalizedAudience":["창업자·스타트업","투자자·VC"],"summary":"이 영상은 AI 경쟁을 단순한 기술 발전이 아니라 미·중 패권 경쟁과 규제 전쟁의 관점에서 해석한다. 특히 DeepSeek의 등장, 중국의 오픈소스 확산, 그리고 미국과 유럽에서 벌어지는 AI 규제 논쟁을 통해 '누가 더 빨리 혁신하느냐'뿐 아니라 '누가 산업의 규칙을 정하느냐'가 중요하다고 말한다.\n\n또한 미국 내부의 주(州)별 AI 법안 난립, 캘리포니아의 SB1047 같은 강경 규제 시도, 그리고 EU AI Act의 부작용을 사례로 들며 과도한 규제가 기술 개발과 오픈소스 생태계를 어떻게 위축시키는지 설명한다. 전반적으로 화자는 미국이 연방 차원에서 주도권을 잡아야 하며, AI는 글로벌 경쟁 속에서 규제의 균형을 잘못 맞추면 자국 산업을 스스로 약화시킬 수 있다고 경고한다.","insights":["AI는 기술 경쟁이면서 동시에 규칙을 정하는 경쟁이다.","과도한 규제는 혁신을 보호하기보다 산업 자체를 질식시킨다.","오픈소스 책임을 과도하게 묻으면 생태계 확산이 멈춘다.","미·중 경쟁 구도는 미국 내 규제 완화의 강한 동력이 된다.","연방과 주 정부의 규제 충돌은 AI 산업에 큰 불확실성을 만든다."],"keyClips":[{"clipId":"xRh2sVcNXQ8:c3:1-8","startSegmentIndex":1,"endSegmentIndex":8,"startTime":1801.99,"endTime":1920.559,"durationSeconds":118.6,"preview":"딥시크가 바꾼 판","mustSee":true},{"clipId":"xRh2sVcNXQ8:c3:9-13","startSegmentIndex":9,"endSegmentIndex":13,"startTime":1920.559,"endTime":1966.48,"durationSeconds":45.9,"preview":"미중 경쟁이 만든 변화","mustSee":false},{"clipId":"xRh2sVcNXQ8:c3:14-19","startSegmentIndex":14,"endSegmentIndex":19,"startTime":1966.48,"endTime":2030.88,"durationSeconds":64.4,"preview":"연방이 먼저다","mustSee":false},{"clipId":"xRh2sVcNXQ8:c3:20-37","startSegmentIndex":20,"endSegmentIndex":37,"startTime":2030.88,"endTime":2191.04,"durationSeconds":160.2,"preview":"주별 규제의 함정","mustSee":false},{"clipId":"xRh2sVcNXQ8:c3:42-60","startSegmentIndex":42,"endSegmentIndex":60,"startTime":2219.52,"endTime":2389.76,"durationSeconds":170.2,"preview":"유럽식 규제의 대가","mustSee":true}],"curatedSegments":[{"segmentIndex":45,"text":"Um, venture We have our issues and venture but a huge advantage that we have is we don't have to we can bet on multiple strategies at the same time right um and we are doing this so we are betting on big models and small models and prepared train models and open source models right and you know and foundation models and applications right uh and consumer and enterprise and so the portfolio approach the nature of it is like we are aggressively basically uh we are aggressively investing behind every strategy that we've identified that we think has a plausible chance of even when that even when that's contradictory to another strategy that we're investing in and one is just like the world's messy and probably a bunch of things are going to work and so like there's not going to be clean yes or no answers to a bunch of this like a lot a answers to a bunch of this like a lot of the answers to this I think are just going to be and answers but the other is like if one of these strategies doesn't work like you know we're not trying to hedge per se but you know we're going to have representation in the portfolio of the alternate strategy and so we're going to have mult multiple ways to win.","startTime":3451.52,"endTime":3506.48,"durationSeconds":55,"level":"C1","overallScore":10,"rationale":"벤처 포트폴리오 전략의 핵심 논리 제시."},{"segmentIndex":34,"text":"Um whereas of course China exports just a tremendous number of physical things right um including like a huge part of like the entire supply chain of parts that basically go into everything that American manufacturers you know kind of make right and so by the time a US you know whatever by the time an American company brings a toy to market right or a uh you know or a car um or anything or a computer or a smartphone or whatever like it's got a lot of componentry in it that was made in China so there is a much tighter in interlinkage between the American and Chinese economies than there as the American and Soviet economies and you know may maybe you know Adam Smith or whatever might say you know that's good news for peace and that you know both countries need each other by the way the other part of that argument is that the Chinese basically the Chinese you know the Chinese governance model is based on high employment um you know because you know if you know at least all the geopolitical people say if China ended up with like 25 or 50% unemployment that would cause civil unrest which is the one thing that the CCP doesn't want and so the corresponding part of the trade pressure is China needs the American export market you know the American consumer is like a third of the global economy.","startTime":1535.52,"endTime":1595.84,"durationSeconds":60,"level":"C2","overallScore":8.8,"rationale":"무역 상호의존과 정치 논리를 깊게 설명."},{"segmentIndex":5,"text":"it came from a hedge fund and it like as far as we can tell it basically is this somewhat idiosyncratic situation where you just have this incredibly successful quant hedge fund with all these you know super geniuses um and the founder of that hedge fund you know basically decided to build AI um and you know at least external indications are this was a surprise to even the Chinese government it's impossible to prove you know what the Chinese government was surprised by or not but you know there's at least the atmospherics are that this was not exactly planned this was not a national champion tech company at the time that Deepseek was released it was it sort of came out of left field which by the way is very encouraging for the field that it was possible for somebody to do that kind of who was unknown right because it kind of means that maybe you don't need all these you know super genius superstar researchers maybe actually smart kids can just build this stuff which I think is the direction things are headed um and so that kicked off I would say like this kind of I don't know copycat's the wrong word but that was sort of it feels like the success of deepseek and the success of deepseek from China as open source kind of kicked off a sort of trend in China releasing these open source models um you know Look, the cynics, you know, in DC would say, you know, yeah, like they're dumping, right?","startTime":1839.12,"endTime":1900.72,"durationSeconds":62,"level":"C1","overallScore":9,"rationale":"배경 해석과 관점, 표현이 매우 풍부."},{"segmentIndex":60,"text":"um you know which is a key way that all this technology proliferates and so this law would have assigned downstream liability to any misuse of open source to the original developer of the open source and so you know you're an independent developer or you're an academic or you're a startup you develop and release an AI model the AI model works fine the day you release it it's great but like 5 years later it gets built into a nuclear power plant and then there's a meltdown of the nuclear","startTime":2379.76,"endTime":2389.76,"durationSeconds":10,"level":"C1","overallScore":9,"rationale":"오픈소스 책임 전가의 문제를 선명히 설명."},{"segmentIndex":27,"text":"Um, and so you've got like the most magic new technology in the world and then it's basically being served up by those companies in a as a cloud business and made basically available to everybody on the planet to just click and use and for like relatively small amounts of money and then on a usage basis which means and usage is great for startups because you it means you can start easily right you the you know there's very you know there's basically no fixed co for a startup building an AI app they don't have giant fixed cost because they could just tap into the open AI or anthropic or Google or Microsoft or whatever you know cloud you know tokens by the you know, intelligence tokens by the drink offering and just get going.","startTime":2607.44,"endTime":2639.839,"durationSeconds":32,"level":"C2","overallScore":9,"rationale":"스타트업 비용구조 통찰과 표현이 풍부."},{"segmentIndex":22,"text":"they're actually they actually as these products get more sophisticated they actually end up using many different kinds of models that are kind of customtailored to the specific aspects of how these products work. Um and so they may start out using one model but they end up using a dozen models and then in the fullness of time it might be 50 or 100 different models for different aspects of the product. Um and so they a lot of these the leading edge application companies are actually backward integrating and actually building their own AI models because they have the deepest understanding of their domain.","startTime":3222.48,"endTime":3234.72,"durationSeconds":12,"level":"C1","overallScore":9,"rationale":"앱 기업의 진화 논리를 깊게 설명."},{"segmentIndex":27,"text":"Um and so that you know that's I think >> small models though right Mark when you think about god models versus small models as you were describing that but that would be small would you categorize that as a small >> well some of them I mean we I will let them announce you know them I will let them announce you know whatever they're doing whenever it's appropriate but some of them are now also doing big model development um and again this is also part of what this is also part of the learning just in the last two years well so like here's a big learning just from the last two years which is very interesting which is two years ago or three years ago for sure you would have said wow open AI is like way out ahead um and like it's probably going to be impossible for anybody to catch up and then it's like okay well Anthropic caught up and so but you know they came out of open AI and so they had all the secrets you know whatever and so knew how to do it and so okay they caught up but surely nobody can catch up after them and then very quickly after that there were a raft of other companies that caught up very fast and XAI is maybe the best example of that which is like you know XAI you know Elon's company XAI is the company name gro is the consumer product version of it um XAI basically caught up to you know state-of-the-art openai anthropic level in like less than 12 months from a standing start right and So, and again that kind of argues against any kind of permanent lead, right, by any one incumbent that's just going to basically be able to lock the entire market down like if you can catch up like that.","startTime":3281.68,"endTime":3291.119,"durationSeconds":9,"level":"C1","overallScore":9,"rationale":"후발 추격 가능성에 대한 핵심 통찰."},{"segmentIndex":13,"text":"I you know my view is we need to be actually very respectful of that and we need to be very aware of that and basically that we you know I use the metaphor with the dog that caught the bus like we always wanted to work on things that matter we are working on things that matter uh people in the rest of society actually really do care about these things um and you know and it's our responsibility to think that all through very carefully and to do a good job um you know both not just building the technology but also explaining it you know look you know I think we have a real obligation to uh you know to really explain ourselves and engage on these issues um in terms of how to measure how going you know it's sort of the classic social science question um uh which is like okay if you want to understand basically you know patterns of people there's basically two ways to understand what people are doing and thinking um one is to ask them and then the other is to watch them um and like every social scientist like every sociologist will tell you this which basically is you can ask people right and the way you do that right is like you know surveys focus groups polls um you know what they think Um but then you can watch them and you then but then you can watch them and you can do what's you know called reveal preferences.","startTime":4316.88,"endTime":4379.28,"durationSeconds":62,"level":"C2","overallScore":9,"rationale":"책임론과 사회과학 틀 제시."},{"segmentIndex":52,"text":"Um and you know you have these like long elaborate you know discussions about you know theories on this and that and the other thing and then reality just like completely smacks you square in the face you know like you idiot right you know like you know what were you like you know this is like the you know the ultimate frustration of the business which is also very motivating which is the number of times that you think that you've applied superior analysis and then you've either invested or not invested based on that analysis and it turns out it was just you the analysis was just completely wrong right um and you know you just like completely overrated your ability to epistemically you know kind of analyze these things you just you know basically inflicted harm like I always the question is always you know it's sort of you know any activity that we do is it value add or is it actually value subtract right and I think in this business of all businesses is kind of like that and that applies to all of my own contributions as well so there is that and then I would say um you know maybe the final thing is just like I do have the entire internet ready to tell me that I'm an idiot so that also doesn't hurt and it does on a regular basis on the point of uh your alluding to earlier about uh decisions on investing in companies.","startTime":4683.36,"endTime":4748.96,"durationSeconds":66,"level":"C2","overallScore":9.2,"rationale":"현실 검증과 오판의 교훈이 강함."},{"segmentIndex":2,"text":"We're seeing companies grow much faster. I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in 5 or 10 years. I think things are going to get much more sophisticated from here.","startTime":8.72,"endTime":16.96,"durationSeconds":8,"level":"C1","overallScore":7.8,"rationale":"제품 진화에 대한 전망이 뚜렷함."},{"segmentIndex":4,"text":"These are trillion dollar questions, not answers. But once somebody proves that it's capable, it seems to not be that hard for other people to be able to catch up, even people with far less resources.","startTime":20.8,"endTime":29.84,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"기술 추격의 본질을 짚는 문장."},{"segmentIndex":9,"text":"One is to ask them and then the other is to watch them. And what you often see in many areas of human activity, including politics and many different aspects of society, the answers that you get when you ask people are very different than the answers that you get when you watch them.","startTime":52.64,"endTime":64.159,"durationSeconds":12,"level":"C1","overallScore":8,"rationale":"말과 행동의 괴리를 통찰함."},{"segmentIndex":27,"text":"Um and so like and that was of course the path that the industry took which was building these kind of hyper literal you know mathematical machines you know that could execute mathematical operations billions of times per second but of course had no ability to kind of deal with human beings the way humans like to be dealt with and so you know couldn't understand you know human speech human language um and so forth and that's the computer industry that got built over the last 80 years and that's the computer industry that's built all the wealth of uh and financial returns of the computer industry uh over the last 80 years you know across all the generations of computers from mainframes through to smartphones.","startTime":174.64,"endTime":206.8,"durationSeconds":32,"level":"C2","overallScore":8,"rationale":"컴퓨터 산업의 한계를 깊게 설명."},{"segmentIndex":34,"text":"Um but and the neural network basically didn't happen but the neural network as an idea continued to be explored in academia um and sort of advanced research by sort of a rump you know movement that was originally called cybernetics and then became known as artificial intelligence uh basically for the last 80 years and essentially it didn't work like essentially it was basically decade after decade of excessive optimism uh followed by disappointment.","startTime":260.639,"endTime":283.199,"durationSeconds":23,"level":"C2","overallScore":8,"rationale":"AI 역사 사이클을 압축해 설명."},{"segmentIndex":41,"text":"So, we're sort of three year we're sort of three years in um to, you know, basically what is effectively an 80-year revolution um of actually being able to deliver on all the promise that the people on the all the on the alternate path, the sort of human cognition model path, you know, kind of saw from the very beginning and then, you know, the great news with this technology is it's already it's kind of ultra democratized.","startTime":325.36,"endTime":347.84000000000003,"durationSeconds":22,"level":"C2","overallScore":8,"rationale":"AI 혁명의 위치와 성격을 압축."},{"segmentIndex":47,"text":"Like I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in five or 10 years.","startTime":423.199,"endTime":426.56,"durationSeconds":3,"level":"C1","overallScore":7.8,"rationale":"미래 제품 변화 전망이 선명함."},{"segmentIndex":15,"text":"Um, and so sort of the internet's the carrier wave for AI to be able to proliferate at kind of light speed uh into the broad base of the global population.","startTime":682.88,"endTime":691.6,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"AI 확산 비유가 강하고 표현도 인상적."},{"segmentIndex":16,"text":"And that's a let's just say that's a potential rate of proliferation of a new technology that's just far faster than has ever been possible before.","startTime":691.6,"endTime":697.8389999999999,"durationSeconds":6,"level":"B2","overallScore":7.4,"rationale":"기술 확산 속도에 대한 통찰이 강함."},{"segmentIndex":19,"text":"Um, you know, you couldn't download television, but you can download AI. Um, and then and an incredible rate. Um, and then they're monetizing really well. Um and the core business model, right, is actually quite interesting. Um and then as a consequence there's kind of this hyperdelation of per unit cost and then that is like driving you know just like you know a more than corresponding level of demand growth you know with elasticity.","startTime":704.8,"endTime":715.2,"durationSeconds":10,"level":"C1","overallScore":8,"rationale":"사업모델·수요 논리가 압축돼 통찰 높음."},{"segmentIndex":25,"text":"Um, and you know, if you have the ability to like inject more intelligence into your business and you have the ability to do, you know, even the most prosaic things like raise your customer service scores, uh, you know, increase upsells, um, uh, you know, or reduce churn or if you have the ability to, um, you know, run marketing campaigns more effectively, um, you know, all of which AI is directly relevant to, like, you know, these are like direct business payoffs, um, you know, that people are seeing already.","startTime":764.48,"endTime":786.639,"durationSeconds":22,"level":"C1","overallScore":8.2,"rationale":"기업 효용 사례와 실전 표현이 매우 풍부함."}],"generatedAt":"2026-06-24T23:20:02.476Z","keyClipsTotalSec":2283},{"videoId":"xRh2sVcNXQ8","chunkIndex":4,"totalChunks":9,"title":"Marc Andreessen's 2026 Outlook: AI Timelines, US vs. China, and The Price of AI — Part 5 of 9","thumbnail":"https://i.ytimg.com/vi/xRh2sVcNXQ8/maxresdefault.jpg","duration":4878,"uploader":"a16z","youtubeUrl":"https://www.youtube.com/watch?v=xRh2sVcNXQ8","keywords":["ai-pricing","open-source-ai","startup","venture-capital","cloud-computing","ai-regulation","business-strategy","technology-policy"],"normalizedKeywords":["비즈니스·전략","기술 트렌드","리더십·매니지먼트"],"targetAudience":[{"who":"창업자","why":"AI 제품의 가격 책정과 시장 구조 변화가 창업 전략에 직결됨"},{"who":"투자자","why":"오픈소스·클로즈드소스와 가격 모델이 수조 원 가치에 영향을 줌"},{"who":"엔지니어","why":"오픈소스 AI가 학습과 개발 진입장벽을 어떻게 낮추는지 이해할 수 있음"},{"who":"정책 관심자","why":"주별 AI 규제가 산업 혁신을 어떻게 위협하는지 알 수 있음"}],"normalizedAudience":["창업자·스타트업","투자자·VC","엔지니어·개발자"],"summary":"이 영상은 AI 산업의 핵심 쟁점을 두 축으로 정리한다. 하나는 AI의 가격을 사용량 기준으로 매기느냐, 가치 기준으로 매기느냐의 문제이고, 다른 하나는 오픈소스와 클로즈드소스 중 무엇이 장기적으로 이길지에 대한 전망이다. 화자는 지금의 클라우드 경쟁 구조 때문에 AI가 빠르게 대중화되고 있으며, 스타트업은 고정비 없이 강력한 모델을 쓸 수 있어 매우 유리하다고 본다.\n\n동시에 그는 AI 가격을 단순히 비용 기준으로 정하면 안 되고, 실제로 창출하는 가치나 생산성 향상에 따라 매겨야 한다고 강조한다. 오픈소스 AI는 경쟁을 촉진할 뿐 아니라 학습과 교육을 가능하게 해 지식 확산을 가속한다는 점도 중요하게 본다. 마지막으로, 과도한 주 규제는 혁신을 막을 수 있으므로 연방 차원의 정리가 필요하다고 주장한다.","insights":["AI 가격은 비용이 아니라 창출가치로 정해야 한다.","사용량 기반 과금은 스타트업의 진입장벽을 크게 낮춘다.","오픈소스는 모델보다 '학습 가능성'에서 더 큰 가치가 있다.","높은 가격과 높은 마진은 오히려 제품 개선을 앞당긴다.","AI 규제는 주별 파편화보다 연방 수준 정리가 필요하다."],"keyClips":[{"clipId":"xRh2sVcNXQ8:c4:1-17","startSegmentIndex":1,"endSegmentIndex":17,"startTime":2400.95,"endTime":2511.92,"durationSeconds":111,"preview":"규제가 혁신을 막는다","mustSee":false},{"clipId":"xRh2sVcNXQ8:c4:19-29","startSegmentIndex":19,"endSegmentIndex":29,"startTime":2515.92,"endTime":2661.92,"durationSeconds":146,"preview":"AI 과금의 본질","mustSee":false},{"clipId":"xRh2sVcNXQ8:c4:30-48","startSegmentIndex":30,"endSegmentIndex":48,"startTime":2661.92,"endTime":2821.44,"durationSeconds":159.5,"preview":"가치기반 가격 실험","mustSee":true},{"clipId":"xRh2sVcNXQ8:c4:50-68","startSegmentIndex":50,"endSegmentIndex":68,"startTime":2830.319,"endTime":2992.48,"durationSeconds":162.2,"preview":"오픈소스의 확산력","mustSee":true}],"curatedSegments":[{"segmentIndex":45,"text":"Um, venture We have our issues and venture but a huge advantage that we have is we don't have to we can bet on multiple strategies at the same time right um and we are doing this so we are betting on big models and small models and prepared train models and open source models right and you know and foundation models and applications right uh and consumer and enterprise and so the portfolio approach the nature of it is like we are aggressively basically uh we are aggressively investing behind every strategy that we've identified that we think has a plausible chance of even when that even when that's contradictory to another strategy that we're investing in and one is just like the world's messy and probably a bunch of things are going to work and so like there's not going to be clean yes or no answers to a bunch of this like a lot a answers to a bunch of this like a lot of the answers to this I think are just going to be and answers but the other is like if one of these strategies doesn't work like you know we're not trying to hedge per se but you know we're going to have representation in the portfolio of the alternate strategy and so we're going to have mult multiple ways to win.","startTime":3451.52,"endTime":3506.48,"durationSeconds":55,"level":"C1","overallScore":10,"rationale":"벤처 포트폴리오 전략의 핵심 논리 제시."},{"segmentIndex":34,"text":"Um whereas of course China exports just a tremendous number of physical things right um including like a huge part of like the entire supply chain of parts that basically go into everything that American manufacturers you know kind of make right and so by the time a US you know whatever by the time an American company brings a toy to market right or a uh you know or a car um or anything or a computer or a smartphone or whatever like it's got a lot of componentry in it that was made in China so there is a much tighter in interlinkage between the American and Chinese economies than there as the American and Soviet economies and you know may maybe you know Adam Smith or whatever might say you know that's good news for peace and that you know both countries need each other by the way the other part of that argument is that the Chinese basically the Chinese you know the Chinese governance model is based on high employment um you know because you know if you know at least all the geopolitical people say if China ended up with like 25 or 50% unemployment that would cause civil unrest which is the one thing that the CCP doesn't want and so the corresponding part of the trade pressure is China needs the American export market you know the American consumer is like a third of the global economy.","startTime":1535.52,"endTime":1595.84,"durationSeconds":60,"level":"C2","overallScore":8.8,"rationale":"무역 상호의존과 정치 논리를 깊게 설명."},{"segmentIndex":5,"text":"it came from a hedge fund and it like as far as we can tell it basically is this somewhat idiosyncratic situation where you just have this incredibly successful quant hedge fund with all these you know super geniuses um and the founder of that hedge fund you know basically decided to build AI um and you know at least external indications are this was a surprise to even the Chinese government it's impossible to prove you know what the Chinese government was surprised by or not but you know there's at least the atmospherics are that this was not exactly planned this was not a national champion tech company at the time that Deepseek was released it was it sort of came out of left field which by the way is very encouraging for the field that it was possible for somebody to do that kind of who was unknown right because it kind of means that maybe you don't need all these you know super genius superstar researchers maybe actually smart kids can just build this stuff which I think is the direction things are headed um and so that kicked off I would say like this kind of I don't know copycat's the wrong word but that was sort of it feels like the success of deepseek and the success of deepseek from China as open source kind of kicked off a sort of trend in China releasing these open source models um you know Look, the cynics, you know, in DC would say, you know, yeah, like they're dumping, right?","startTime":1839.12,"endTime":1900.72,"durationSeconds":62,"level":"C1","overallScore":9,"rationale":"배경 해석과 관점, 표현이 매우 풍부."},{"segmentIndex":60,"text":"um you know which is a key way that all this technology proliferates and so this law would have assigned downstream liability to any misuse of open source to the original developer of the open source and so you know you're an independent developer or you're an academic or you're a startup you develop and release an AI model the AI model works fine the day you release it it's great but like 5 years later it gets built into a nuclear power plant and then there's a meltdown of the nuclear","startTime":2379.76,"endTime":2389.76,"durationSeconds":10,"level":"C1","overallScore":9,"rationale":"오픈소스 책임 전가의 문제를 선명히 설명."},{"segmentIndex":27,"text":"Um, and so you've got like the most magic new technology in the world and then it's basically being served up by those companies in a as a cloud business and made basically available to everybody on the planet to just click and use and for like relatively small amounts of money and then on a usage basis which means and usage is great for startups because you it means you can start easily right you the you know there's very you know there's basically no fixed co for a startup building an AI app they don't have giant fixed cost because they could just tap into the open AI or anthropic or Google or Microsoft or whatever you know cloud you know tokens by the you know, intelligence tokens by the drink offering and just get going.","startTime":2607.44,"endTime":2639.839,"durationSeconds":32,"level":"C2","overallScore":9,"rationale":"스타트업 비용구조 통찰과 표현이 풍부."},{"segmentIndex":22,"text":"they're actually they actually as these products get more sophisticated they actually end up using many different kinds of models that are kind of customtailored to the specific aspects of how these products work. Um and so they may start out using one model but they end up using a dozen models and then in the fullness of time it might be 50 or 100 different models for different aspects of the product. Um and so they a lot of these the leading edge application companies are actually backward integrating and actually building their own AI models because they have the deepest understanding of their domain.","startTime":3222.48,"endTime":3234.72,"durationSeconds":12,"level":"C1","overallScore":9,"rationale":"앱 기업의 진화 논리를 깊게 설명."},{"segmentIndex":27,"text":"Um and so that you know that's I think >> small models though right Mark when you think about god models versus small models as you were describing that but that would be small would you categorize that as a small >> well some of them I mean we I will let them announce you know them I will let them announce you know whatever they're doing whenever it's appropriate but some of them are now also doing big model development um and again this is also part of what this is also part of the learning just in the last two years well so like here's a big learning just from the last two years which is very interesting which is two years ago or three years ago for sure you would have said wow open AI is like way out ahead um and like it's probably going to be impossible for anybody to catch up and then it's like okay well Anthropic caught up and so but you know they came out of open AI and so they had all the secrets you know whatever and so knew how to do it and so okay they caught up but surely nobody can catch up after them and then very quickly after that there were a raft of other companies that caught up very fast and XAI is maybe the best example of that which is like you know XAI you know Elon's company XAI is the company name gro is the consumer product version of it um XAI basically caught up to you know state-of-the-art openai anthropic level in like less than 12 months from a standing start right and So, and again that kind of argues against any kind of permanent lead, right, by any one incumbent that's just going to basically be able to lock the entire market down like if you can catch up like that.","startTime":3281.68,"endTime":3291.119,"durationSeconds":9,"level":"C1","overallScore":9,"rationale":"후발 추격 가능성에 대한 핵심 통찰."},{"segmentIndex":13,"text":"I you know my view is we need to be actually very respectful of that and we need to be very aware of that and basically that we you know I use the metaphor with the dog that caught the bus like we always wanted to work on things that matter we are working on things that matter uh people in the rest of society actually really do care about these things um and you know and it's our responsibility to think that all through very carefully and to do a good job um you know both not just building the technology but also explaining it you know look you know I think we have a real obligation to uh you know to really explain ourselves and engage on these issues um in terms of how to measure how going you know it's sort of the classic social science question um uh which is like okay if you want to understand basically you know patterns of people there's basically two ways to understand what people are doing and thinking um one is to ask them and then the other is to watch them um and like every social scientist like every sociologist will tell you this which basically is you can ask people right and the way you do that right is like you know surveys focus groups polls um you know what they think Um but then you can watch them and you then but then you can watch them and you can do what's you know called reveal preferences.","startTime":4316.88,"endTime":4379.28,"durationSeconds":62,"level":"C2","overallScore":9,"rationale":"책임론과 사회과학 틀 제시."},{"segmentIndex":52,"text":"Um and you know you have these like long elaborate you know discussions about you know theories on this and that and the other thing and then reality just like completely smacks you square in the face you know like you idiot right you know like you know what were you like you know this is like the you know the ultimate frustration of the business which is also very motivating which is the number of times that you think that you've applied superior analysis and then you've either invested or not invested based on that analysis and it turns out it was just you the analysis was just completely wrong right um and you know you just like completely overrated your ability to epistemically you know kind of analyze these things you just you know basically inflicted harm like I always the question is always you know it's sort of you know any activity that we do is it value add or is it actually value subtract right and I think in this business of all businesses is kind of like that and that applies to all of my own contributions as well so there is that and then I would say um you know maybe the final thing is just like I do have the entire internet ready to tell me that I'm an idiot so that also doesn't hurt and it does on a regular basis on the point of uh your alluding to earlier about uh decisions on investing in companies.","startTime":4683.36,"endTime":4748.96,"durationSeconds":66,"level":"C2","overallScore":9.2,"rationale":"현실 검증과 오판의 교훈이 강함."},{"segmentIndex":2,"text":"We're seeing companies grow much faster. I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in 5 or 10 years. I think things are going to get much more sophisticated from here.","startTime":8.72,"endTime":16.96,"durationSeconds":8,"level":"C1","overallScore":7.8,"rationale":"제품 진화에 대한 전망이 뚜렷함."},{"segmentIndex":4,"text":"These are trillion dollar questions, not answers. But once somebody proves that it's capable, it seems to not be that hard for other people to be able to catch up, even people with far less resources.","startTime":20.8,"endTime":29.84,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"기술 추격의 본질을 짚는 문장."},{"segmentIndex":9,"text":"One is to ask them and then the other is to watch them. And what you often see in many areas of human activity, including politics and many different aspects of society, the answers that you get when you ask people are very different than the answers that you get when you watch them.","startTime":52.64,"endTime":64.159,"durationSeconds":12,"level":"C1","overallScore":8,"rationale":"말과 행동의 괴리를 통찰함."},{"segmentIndex":27,"text":"Um and so like and that was of course the path that the industry took which was building these kind of hyper literal you know mathematical machines you know that could execute mathematical operations billions of times per second but of course had no ability to kind of deal with human beings the way humans like to be dealt with and so you know couldn't understand you know human speech human language um and so forth and that's the computer industry that got built over the last 80 years and that's the computer industry that's built all the wealth of uh and financial returns of the computer industry uh over the last 80 years you know across all the generations of computers from mainframes through to smartphones.","startTime":174.64,"endTime":206.8,"durationSeconds":32,"level":"C2","overallScore":8,"rationale":"컴퓨터 산업의 한계를 깊게 설명."},{"segmentIndex":34,"text":"Um but and the neural network basically didn't happen but the neural network as an idea continued to be explored in academia um and sort of advanced research by sort of a rump you know movement that was originally called cybernetics and then became known as artificial intelligence uh basically for the last 80 years and essentially it didn't work like essentially it was basically decade after decade of excessive optimism uh followed by disappointment.","startTime":260.639,"endTime":283.199,"durationSeconds":23,"level":"C2","overallScore":8,"rationale":"AI 역사 사이클을 압축해 설명."},{"segmentIndex":41,"text":"So, we're sort of three year we're sort of three years in um to, you know, basically what is effectively an 80-year revolution um of actually being able to deliver on all the promise that the people on the all the on the alternate path, the sort of human cognition model path, you know, kind of saw from the very beginning and then, you know, the great news with this technology is it's already it's kind of ultra democratized.","startTime":325.36,"endTime":347.84000000000003,"durationSeconds":22,"level":"C2","overallScore":8,"rationale":"AI 혁명의 위치와 성격을 압축."},{"segmentIndex":47,"text":"Like I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in five or 10 years.","startTime":423.199,"endTime":426.56,"durationSeconds":3,"level":"C1","overallScore":7.8,"rationale":"미래 제품 변화 전망이 선명함."},{"segmentIndex":15,"text":"Um, and so sort of the internet's the carrier wave for AI to be able to proliferate at kind of light speed uh into the broad base of the global population.","startTime":682.88,"endTime":691.6,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"AI 확산 비유가 강하고 표현도 인상적."},{"segmentIndex":16,"text":"And that's a let's just say that's a potential rate of proliferation of a new technology that's just far faster than has ever been possible before.","startTime":691.6,"endTime":697.8389999999999,"durationSeconds":6,"level":"B2","overallScore":7.4,"rationale":"기술 확산 속도에 대한 통찰이 강함."},{"segmentIndex":19,"text":"Um, you know, you couldn't download television, but you can download AI. Um, and then and an incredible rate. Um, and then they're monetizing really well. Um and the core business model, right, is actually quite interesting. Um and then as a consequence there's kind of this hyperdelation of per unit cost and then that is like driving you know just like you know a more than corresponding level of demand growth you know with elasticity.","startTime":704.8,"endTime":715.2,"durationSeconds":10,"level":"C1","overallScore":8,"rationale":"사업모델·수요 논리가 압축돼 통찰 높음."},{"segmentIndex":25,"text":"Um, and you know, if you have the ability to like inject more intelligence into your business and you have the ability to do, you know, even the most prosaic things like raise your customer service scores, uh, you know, increase upsells, um, uh, you know, or reduce churn or if you have the ability to, um, you know, run marketing campaigns more effectively, um, you know, all of which AI is directly relevant to, like, you know, these are like direct business payoffs, um, you know, that people are seeing already.","startTime":764.48,"endTime":786.639,"durationSeconds":22,"level":"C1","overallScore":8.2,"rationale":"기업 효용 사례와 실전 표현이 매우 풍부함."}],"generatedAt":"2026-06-24T23:20:23.581Z","keyClipsTotalSec":2283},{"videoId":"xRh2sVcNXQ8","chunkIndex":5,"totalChunks":9,"title":"Marc Andreessen's 2026 Outlook: AI Timelines, US vs. China, and The Price of AI — Part 6 of 9","thumbnail":"https://i.ytimg.com/vi/xRh2sVcNXQ8/maxresdefault.jpg","duration":4878,"uploader":"a16z","youtubeUrl":"https://www.youtube.com/watch?v=xRh2sVcNXQ8","keywords":["ai","startup","venture-capital","foundation-models","generative-ai","market-dynamics","china","open-source","enterprise-software","tech-industry"],"normalizedKeywords":["비즈니스·전략","기술 트렌드","프로덕트"],"targetAudience":[{"who":"창업자","why":"AI 시장의 승자 구조와 제품 전략을 읽는 데 도움이 됨"},{"who":"투자자","why":"기반모델·응용층·중국 경쟁의 판세를 포트폴리오 관점에서 볼 수 있음"},{"who":"엔지니어","why":"모델 조합, 백워드 통합, 오픈소스 활용 흐름을 이해하는 데 유용함"}],"normalizedAudience":["창업자·스타트업","투자자·VC","엔지니어·개발자"],"summary":"이 영상은 AI 산업에서 누가 이기고 지는지, 그리고 그 판이 왜 생각보다 훨씬 더 빠르게 바뀌는지를 다룬다. 마크 안드레센은 오픈AI 같은 선두주자도 영원한 독점 우위를 갖기 어렵고, Anthropic·xAI·중국 기업들처럼 새로운 경쟁자들이 짧은 시간에 따라붙을 수 있다고 본다. 그 결과 단일 승자보다 고성능 대형 모델과 저가의 소형 모델이 공존하는 구조, 그리고 기반모델·오픈소스·응용회사의 다중 전략이 동시에 살아남는 구조를 강조한다.\n\n특히 그는 'GPT 래퍼'라고 비판받던 AI 응용회사가 사실은 여러 모델을 조합하고, 결국 자기 도메인 이해를 바탕으로 자체 모델까지 만드는 방향으로 진화하고 있다고 말한다. 또 벤처는 정답 하나에 베팅하는 회사가 아니라 여러 전략에 동시에 투자할 수 있기 때문에, 불확실성 자체가 오히려 유리하다고 주장한다. 후반부에는 미국과 중국의 경쟁이 오히려 미국 시스템을 자극하는 긍정적 압력이라는 관점도 제시한다.","insights":["AI 시장은 아직 승자독식이 아니라 추격전이다.","한번 가능성이 증명되면 추격 속도는 매우 빨라진다.","응용회사는 단순 래퍼가 아니라 모델 회사로 커진다.","대형 모델과 소형 모델은 대체재보다 공존재에 가깝다.","불확실성이 클수록 포트폴리오 전략의 가치가 커진다."],"keyClips":[{"clipId":"xRh2sVcNXQ8:c5:4-6","startSegmentIndex":4,"endSegmentIndex":6,"startTime":3018.72,"endTime":3039.599,"durationSeconds":20.9,"preview":"대형과 소형 공존","mustSee":false},{"clipId":"xRh2sVcNXQ8:c5:10-15","startSegmentIndex":10,"endSegmentIndex":15,"startTime":3068.16,"endTime":3136.079,"durationSeconds":67.9,"preview":"새 인커밴트의 등장","mustSee":true},{"clipId":"xRh2sVcNXQ8:c5:18-26","startSegmentIndex":18,"endSegmentIndex":26,"startTime":3165.599,"endTime":3281.68,"durationSeconds":116.1,"preview":"응용회사의 역진화","mustSee":false},{"clipId":"xRh2sVcNXQ8:c5:29-33","startSegmentIndex":29,"endSegmentIndex":33,"startTime":3350.16,"endTime":3383.04,"durationSeconds":32.9,"preview":"중국 추격의 의미","mustSee":false},{"clipId":"xRh2sVcNXQ8:c5:42-47","startSegmentIndex":42,"endSegmentIndex":47,"startTime":3423.92,"endTime":3519.28,"durationSeconds":95.4,"preview":"벡터는 포트폴리오","mustSee":false}],"curatedSegments":[{"segmentIndex":45,"text":"Um, venture We have our issues and venture but a huge advantage that we have is we don't have to we can bet on multiple strategies at the same time right um and we are doing this so we are betting on big models and small models and prepared train models and open source models right and you know and foundation models and applications right uh and consumer and enterprise and so the portfolio approach the nature of it is like we are aggressively basically uh we are aggressively investing behind every strategy that we've identified that we think has a plausible chance of even when that even when that's contradictory to another strategy that we're investing in and one is just like the world's messy and probably a bunch of things are going to work and so like there's not going to be clean yes or no answers to a bunch of this like a lot a answers to a bunch of this like a lot of the answers to this I think are just going to be and answers but the other is like if one of these strategies doesn't work like you know we're not trying to hedge per se but you know we're going to have representation in the portfolio of the alternate strategy and so we're going to have mult multiple ways to win.","startTime":3451.52,"endTime":3506.48,"durationSeconds":55,"level":"C1","overallScore":10,"rationale":"벤처 포트폴리오 전략의 핵심 논리 제시."},{"segmentIndex":34,"text":"Um whereas of course China exports just a tremendous number of physical things right um including like a huge part of like the entire supply chain of parts that basically go into everything that American manufacturers you know kind of make right and so by the time a US you know whatever by the time an American company brings a toy to market right or a uh you know or a car um or anything or a computer or a smartphone or whatever like it's got a lot of componentry in it that was made in China so there is a much tighter in interlinkage between the American and Chinese economies than there as the American and Soviet economies and you know may maybe you know Adam Smith or whatever might say you know that's good news for peace and that you know both countries need each other by the way the other part of that argument is that the Chinese basically the Chinese you know the Chinese governance model is based on high employment um you know because you know if you know at least all the geopolitical people say if China ended up with like 25 or 50% unemployment that would cause civil unrest which is the one thing that the CCP doesn't want and so the corresponding part of the trade pressure is China needs the American export market you know the American consumer is like a third of the global economy.","startTime":1535.52,"endTime":1595.84,"durationSeconds":60,"level":"C2","overallScore":8.8,"rationale":"무역 상호의존과 정치 논리를 깊게 설명."},{"segmentIndex":5,"text":"it came from a hedge fund and it like as far as we can tell it basically is this somewhat idiosyncratic situation where you just have this incredibly successful quant hedge fund with all these you know super geniuses um and the founder of that hedge fund you know basically decided to build AI um and you know at least external indications are this was a surprise to even the Chinese government it's impossible to prove you know what the Chinese government was surprised by or not but you know there's at least the atmospherics are that this was not exactly planned this was not a national champion tech company at the time that Deepseek was released it was it sort of came out of left field which by the way is very encouraging for the field that it was possible for somebody to do that kind of who was unknown right because it kind of means that maybe you don't need all these you know super genius superstar researchers maybe actually smart kids can just build this stuff which I think is the direction things are headed um and so that kicked off I would say like this kind of I don't know copycat's the wrong word but that was sort of it feels like the success of deepseek and the success of deepseek from China as open source kind of kicked off a sort of trend in China releasing these open source models um you know Look, the cynics, you know, in DC would say, you know, yeah, like they're dumping, right?","startTime":1839.12,"endTime":1900.72,"durationSeconds":62,"level":"C1","overallScore":9,"rationale":"배경 해석과 관점, 표현이 매우 풍부."},{"segmentIndex":60,"text":"um you know which is a key way that all this technology proliferates and so this law would have assigned downstream liability to any misuse of open source to the original developer of the open source and so you know you're an independent developer or you're an academic or you're a startup you develop and release an AI model the AI model works fine the day you release it it's great but like 5 years later it gets built into a nuclear power plant and then there's a meltdown of the nuclear","startTime":2379.76,"endTime":2389.76,"durationSeconds":10,"level":"C1","overallScore":9,"rationale":"오픈소스 책임 전가의 문제를 선명히 설명."},{"segmentIndex":27,"text":"Um, and so you've got like the most magic new technology in the world and then it's basically being served up by those companies in a as a cloud business and made basically available to everybody on the planet to just click and use and for like relatively small amounts of money and then on a usage basis which means and usage is great for startups because you it means you can start easily right you the you know there's very you know there's basically no fixed co for a startup building an AI app they don't have giant fixed cost because they could just tap into the open AI or anthropic or Google or Microsoft or whatever you know cloud you know tokens by the you know, intelligence tokens by the drink offering and just get going.","startTime":2607.44,"endTime":2639.839,"durationSeconds":32,"level":"C2","overallScore":9,"rationale":"스타트업 비용구조 통찰과 표현이 풍부."},{"segmentIndex":22,"text":"they're actually they actually as these products get more sophisticated they actually end up using many different kinds of models that are kind of customtailored to the specific aspects of how these products work. Um and so they may start out using one model but they end up using a dozen models and then in the fullness of time it might be 50 or 100 different models for different aspects of the product. Um and so they a lot of these the leading edge application companies are actually backward integrating and actually building their own AI models because they have the deepest understanding of their domain.","startTime":3222.48,"endTime":3234.72,"durationSeconds":12,"level":"C1","overallScore":9,"rationale":"앱 기업의 진화 논리를 깊게 설명."},{"segmentIndex":27,"text":"Um and so that you know that's I think >> small models though right Mark when you think about god models versus small models as you were describing that but that would be small would you categorize that as a small >> well some of them I mean we I will let them announce you know them I will let them announce you know whatever they're doing whenever it's appropriate but some of them are now also doing big model development um and again this is also part of what this is also part of the learning just in the last two years well so like here's a big learning just from the last two years which is very interesting which is two years ago or three years ago for sure you would have said wow open AI is like way out ahead um and like it's probably going to be impossible for anybody to catch up and then it's like okay well Anthropic caught up and so but you know they came out of open AI and so they had all the secrets you know whatever and so knew how to do it and so okay they caught up but surely nobody can catch up after them and then very quickly after that there were a raft of other companies that caught up very fast and XAI is maybe the best example of that which is like you know XAI you know Elon's company XAI is the company name gro is the consumer product version of it um XAI basically caught up to you know state-of-the-art openai anthropic level in like less than 12 months from a standing start right and So, and again that kind of argues against any kind of permanent lead, right, by any one incumbent that's just going to basically be able to lock the entire market down like if you can catch up like that.","startTime":3281.68,"endTime":3291.119,"durationSeconds":9,"level":"C1","overallScore":9,"rationale":"후발 추격 가능성에 대한 핵심 통찰."},{"segmentIndex":13,"text":"I you know my view is we need to be actually very respectful of that and we need to be very aware of that and basically that we you know I use the metaphor with the dog that caught the bus like we always wanted to work on things that matter we are working on things that matter uh people in the rest of society actually really do care about these things um and you know and it's our responsibility to think that all through very carefully and to do a good job um you know both not just building the technology but also explaining it you know look you know I think we have a real obligation to uh you know to really explain ourselves and engage on these issues um in terms of how to measure how going you know it's sort of the classic social science question um uh which is like okay if you want to understand basically you know patterns of people there's basically two ways to understand what people are doing and thinking um one is to ask them and then the other is to watch them um and like every social scientist like every sociologist will tell you this which basically is you can ask people right and the way you do that right is like you know surveys focus groups polls um you know what they think Um but then you can watch them and you then but then you can watch them and you can do what's you know called reveal preferences.","startTime":4316.88,"endTime":4379.28,"durationSeconds":62,"level":"C2","overallScore":9,"rationale":"책임론과 사회과학 틀 제시."},{"segmentIndex":52,"text":"Um and you know you have these like long elaborate you know discussions about you know theories on this and that and the other thing and then reality just like completely smacks you square in the face you know like you idiot right you know like you know what were you like you know this is like the you know the ultimate frustration of the business which is also very motivating which is the number of times that you think that you've applied superior analysis and then you've either invested or not invested based on that analysis and it turns out it was just you the analysis was just completely wrong right um and you know you just like completely overrated your ability to epistemically you know kind of analyze these things you just you know basically inflicted harm like I always the question is always you know it's sort of you know any activity that we do is it value add or is it actually value subtract right and I think in this business of all businesses is kind of like that and that applies to all of my own contributions as well so there is that and then I would say um you know maybe the final thing is just like I do have the entire internet ready to tell me that I'm an idiot so that also doesn't hurt and it does on a regular basis on the point of uh your alluding to earlier about uh decisions on investing in companies.","startTime":4683.36,"endTime":4748.96,"durationSeconds":66,"level":"C2","overallScore":9.2,"rationale":"현실 검증과 오판의 교훈이 강함."},{"segmentIndex":2,"text":"We're seeing companies grow much faster. I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in 5 or 10 years. I think things are going to get much more sophisticated from here.","startTime":8.72,"endTime":16.96,"durationSeconds":8,"level":"C1","overallScore":7.8,"rationale":"제품 진화에 대한 전망이 뚜렷함."},{"segmentIndex":4,"text":"These are trillion dollar questions, not answers. But once somebody proves that it's capable, it seems to not be that hard for other people to be able to catch up, even people with far less resources.","startTime":20.8,"endTime":29.84,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"기술 추격의 본질을 짚는 문장."},{"segmentIndex":9,"text":"One is to ask them and then the other is to watch them. And what you often see in many areas of human activity, including politics and many different aspects of society, the answers that you get when you ask people are very different than the answers that you get when you watch them.","startTime":52.64,"endTime":64.159,"durationSeconds":12,"level":"C1","overallScore":8,"rationale":"말과 행동의 괴리를 통찰함."},{"segmentIndex":27,"text":"Um and so like and that was of course the path that the industry took which was building these kind of hyper literal you know mathematical machines you know that could execute mathematical operations billions of times per second but of course had no ability to kind of deal with human beings the way humans like to be dealt with and so you know couldn't understand you know human speech human language um and so forth and that's the computer industry that got built over the last 80 years and that's the computer industry that's built all the wealth of uh and financial returns of the computer industry uh over the last 80 years you know across all the generations of computers from mainframes through to smartphones.","startTime":174.64,"endTime":206.8,"durationSeconds":32,"level":"C2","overallScore":8,"rationale":"컴퓨터 산업의 한계를 깊게 설명."},{"segmentIndex":34,"text":"Um but and the neural network basically didn't happen but the neural network as an idea continued to be explored in academia um and sort of advanced research by sort of a rump you know movement that was originally called cybernetics and then became known as artificial intelligence uh basically for the last 80 years and essentially it didn't work like essentially it was basically decade after decade of excessive optimism uh followed by disappointment.","startTime":260.639,"endTime":283.199,"durationSeconds":23,"level":"C2","overallScore":8,"rationale":"AI 역사 사이클을 압축해 설명."},{"segmentIndex":41,"text":"So, we're sort of three year we're sort of three years in um to, you know, basically what is effectively an 80-year revolution um of actually being able to deliver on all the promise that the people on the all the on the alternate path, the sort of human cognition model path, you know, kind of saw from the very beginning and then, you know, the great news with this technology is it's already it's kind of ultra democratized.","startTime":325.36,"endTime":347.84000000000003,"durationSeconds":22,"level":"C2","overallScore":8,"rationale":"AI 혁명의 위치와 성격을 압축."},{"segmentIndex":47,"text":"Like I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in five or 10 years.","startTime":423.199,"endTime":426.56,"durationSeconds":3,"level":"C1","overallScore":7.8,"rationale":"미래 제품 변화 전망이 선명함."},{"segmentIndex":15,"text":"Um, and so sort of the internet's the carrier wave for AI to be able to proliferate at kind of light speed uh into the broad base of the global population.","startTime":682.88,"endTime":691.6,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"AI 확산 비유가 강하고 표현도 인상적."},{"segmentIndex":16,"text":"And that's a let's just say that's a potential rate of proliferation of a new technology that's just far faster than has ever been possible before.","startTime":691.6,"endTime":697.8389999999999,"durationSeconds":6,"level":"B2","overallScore":7.4,"rationale":"기술 확산 속도에 대한 통찰이 강함."},{"segmentIndex":19,"text":"Um, you know, you couldn't download television, but you can download AI. Um, and then and an incredible rate. Um, and then they're monetizing really well. Um and the core business model, right, is actually quite interesting. Um and then as a consequence there's kind of this hyperdelation of per unit cost and then that is like driving you know just like you know a more than corresponding level of demand growth you know with elasticity.","startTime":704.8,"endTime":715.2,"durationSeconds":10,"level":"C1","overallScore":8,"rationale":"사업모델·수요 논리가 압축돼 통찰 높음."},{"segmentIndex":25,"text":"Um, and you know, if you have the ability to like inject more intelligence into your business and you have the ability to do, you know, even the most prosaic things like raise your customer service scores, uh, you know, increase upsells, um, uh, you know, or reduce churn or if you have the ability to, um, you know, run marketing campaigns more effectively, um, you know, all of which AI is directly relevant to, like, you know, these are like direct business payoffs, um, you know, that people are seeing already.","startTime":764.48,"endTime":786.639,"durationSeconds":22,"level":"C1","overallScore":8.2,"rationale":"기업 효용 사례와 실전 표현이 매우 풍부함."}],"generatedAt":"2026-06-24T23:20:58.409Z","keyClipsTotalSec":2283},{"videoId":"xRh2sVcNXQ8","chunkIndex":6,"totalChunks":9,"title":"Marc Andreessen's 2026 Outlook: AI Timelines, US vs. China, and The Price of AI — Part 7 of 9","thumbnail":"https://i.ytimg.com/vi/xRh2sVcNXQ8/maxresdefault.jpg","duration":4878,"uploader":"a16z","youtubeUrl":"https://www.youtube.com/watch?v=xRh2sVcNXQ8","keywords":["venture-capital","ai","startup","technology-trends","public-policy","crypto","biotech","market-strategy","silicon-valley"],"normalizedKeywords":["비즈니스·전략","기술 트렌드","리더십·매니지먼트"],"targetAudience":[{"who":"투자자","why":"기술 파도에 맞춰 투자 주제를 재편하는 VC의 판단 기준을 볼 수 있음"},{"who":"창업자","why":"AI·크립토 같은 구조적 변화기에 어떤 신호를 읽어야 하는지 배울 수 있음"},{"who":"정책 관심자","why":"기술 업계가 왜 워싱턴을 상대로 메시지를 설계하는지 이해할 수 있음"}],"normalizedAudience":["투자자·VC","창업자·스타트업"],"summary":"이 영상은 안드리슨 호로위츠가 왜 AI를 중심축으로 재편했는지, 그리고 벤처투자가 어떤 원리로 성공하는지를 설명한다. 핵심 주장은 벤처의 본질은 '기술의 구조적 전환'에 올라타는 것이며, 인터넷·크립토·AI처럼 새로운 파도가 올 때 적극적으로 이동하지 않으면 생존하기 어렵다는 것이다. 또한 회사의 공개 발언과 논쟁적 포지셔닝은 외부 논란을 만들 수 있지만, 최고의 창업자들에게는 오히려 정체성과 실력을 미리 보여주는 경쟁우위가 된다고 말한다.\n\n후반부에서는 AI가 AD(관련 투자 영역) 전반, 특히 에너지·데이터센터·소재·바이오/헬스케어에 어떤 수요와 결합을 만들어내는지, 그리고 정책 환경 변화로 크립토가 다시 유망해졌다는 점을 짚는다. 전반적으로 이 대담은 '새 기술이 등장할 때 어떤 투자자와 회사가 살아남는가'를 보여주는 전략적 인터뷰다.","insights":["벤처의 본질은 새 기술 파도에 올라타는 것이다.","과감한 공개 메시지는 논란보다 신뢰를 더 빨리 만든다.","창업자는 베일보다 정체성의 선명함에 끌린다.","AI는 단일 산업이 아니라 여러 섹터의 수요를 동시에 바꾼다.","기술 트렌드를 외면하는 VC는 결국 다음 세대에서 밀린다."],"keyClips":[{"clipId":"xRh2sVcNXQ8:c6:1-24","startSegmentIndex":1,"endSegmentIndex":24,"startTime":3600.47,"endTime":3783.2,"durationSeconds":182.7,"preview":"공개 발언의 전략","mustSee":false},{"clipId":"xRh2sVcNXQ8:c6:29-52","startSegmentIndex":29,"endSegmentIndex":52,"startTime":3818.88,"endTime":4032,"durationSeconds":213.1,"preview":"AI에 올라타는 법","mustSee":true},{"clipId":"xRh2sVcNXQ8:c6:53-57","startSegmentIndex":53,"endSegmentIndex":57,"startTime":4032,"endTime":4080.88,"durationSeconds":48.9,"preview":"AI의 확장효과","mustSee":false},{"clipId":"xRh2sVcNXQ8:c6:61-67","startSegmentIndex":61,"endSegmentIndex":67,"startTime":4124.319,"endTime":4209.92,"durationSeconds":85.6,"preview":"사회적 수용의 속도","mustSee":false}],"curatedSegments":[{"segmentIndex":45,"text":"Um, venture We have our issues and venture but a huge advantage that we have is we don't have to we can bet on multiple strategies at the same time right um and we are doing this so we are betting on big models and small models and prepared train models and open source models right and you know and foundation models and applications right uh and consumer and enterprise and so the portfolio approach the nature of it is like we are aggressively basically uh we are aggressively investing behind every strategy that we've identified that we think has a plausible chance of even when that even when that's contradictory to another strategy that we're investing in and one is just like the world's messy and probably a bunch of things are going to work and so like there's not going to be clean yes or no answers to a bunch of this like a lot a answers to a bunch of this like a lot of the answers to this I think are just going to be and answers but the other is like if one of these strategies doesn't work like you know we're not trying to hedge per se but you know we're going to have representation in the portfolio of the alternate strategy and so we're going to have mult multiple ways to win.","startTime":3451.52,"endTime":3506.48,"durationSeconds":55,"level":"C1","overallScore":10,"rationale":"벤처 포트폴리오 전략의 핵심 논리 제시."},{"segmentIndex":34,"text":"Um whereas of course China exports just a tremendous number of physical things right um including like a huge part of like the entire supply chain of parts that basically go into everything that American manufacturers you know kind of make right and so by the time a US you know whatever by the time an American company brings a toy to market right or a uh you know or a car um or anything or a computer or a smartphone or whatever like it's got a lot of componentry in it that was made in China so there is a much tighter in interlinkage between the American and Chinese economies than there as the American and Soviet economies and you know may maybe you know Adam Smith or whatever might say you know that's good news for peace and that you know both countries need each other by the way the other part of that argument is that the Chinese basically the Chinese you know the Chinese governance model is based on high employment um you know because you know if you know at least all the geopolitical people say if China ended up with like 25 or 50% unemployment that would cause civil unrest which is the one thing that the CCP doesn't want and so the corresponding part of the trade pressure is China needs the American export market you know the American consumer is like a third of the global economy.","startTime":1535.52,"endTime":1595.84,"durationSeconds":60,"level":"C2","overallScore":8.8,"rationale":"무역 상호의존과 정치 논리를 깊게 설명."},{"segmentIndex":5,"text":"it came from a hedge fund and it like as far as we can tell it basically is this somewhat idiosyncratic situation where you just have this incredibly successful quant hedge fund with all these you know super geniuses um and the founder of that hedge fund you know basically decided to build AI um and you know at least external indications are this was a surprise to even the Chinese government it's impossible to prove you know what the Chinese government was surprised by or not but you know there's at least the atmospherics are that this was not exactly planned this was not a national champion tech company at the time that Deepseek was released it was it sort of came out of left field which by the way is very encouraging for the field that it was possible for somebody to do that kind of who was unknown right because it kind of means that maybe you don't need all these you know super genius superstar researchers maybe actually smart kids can just build this stuff which I think is the direction things are headed um and so that kicked off I would say like this kind of I don't know copycat's the wrong word but that was sort of it feels like the success of deepseek and the success of deepseek from China as open source kind of kicked off a sort of trend in China releasing these open source models um you know Look, the cynics, you know, in DC would say, you know, yeah, like they're dumping, right?","startTime":1839.12,"endTime":1900.72,"durationSeconds":62,"level":"C1","overallScore":9,"rationale":"배경 해석과 관점, 표현이 매우 풍부."},{"segmentIndex":60,"text":"um you know which is a key way that all this technology proliferates and so this law would have assigned downstream liability to any misuse of open source to the original developer of the open source and so you know you're an independent developer or you're an academic or you're a startup you develop and release an AI model the AI model works fine the day you release it it's great but like 5 years later it gets built into a nuclear power plant and then there's a meltdown of the nuclear","startTime":2379.76,"endTime":2389.76,"durationSeconds":10,"level":"C1","overallScore":9,"rationale":"오픈소스 책임 전가의 문제를 선명히 설명."},{"segmentIndex":27,"text":"Um, and so you've got like the most magic new technology in the world and then it's basically being served up by those companies in a as a cloud business and made basically available to everybody on the planet to just click and use and for like relatively small amounts of money and then on a usage basis which means and usage is great for startups because you it means you can start easily right you the you know there's very you know there's basically no fixed co for a startup building an AI app they don't have giant fixed cost because they could just tap into the open AI or anthropic or Google or Microsoft or whatever you know cloud you know tokens by the you know, intelligence tokens by the drink offering and just get going.","startTime":2607.44,"endTime":2639.839,"durationSeconds":32,"level":"C2","overallScore":9,"rationale":"스타트업 비용구조 통찰과 표현이 풍부."},{"segmentIndex":22,"text":"they're actually they actually as these products get more sophisticated they actually end up using many different kinds of models that are kind of customtailored to the specific aspects of how these products work. Um and so they may start out using one model but they end up using a dozen models and then in the fullness of time it might be 50 or 100 different models for different aspects of the product. Um and so they a lot of these the leading edge application companies are actually backward integrating and actually building their own AI models because they have the deepest understanding of their domain.","startTime":3222.48,"endTime":3234.72,"durationSeconds":12,"level":"C1","overallScore":9,"rationale":"앱 기업의 진화 논리를 깊게 설명."},{"segmentIndex":27,"text":"Um and so that you know that's I think >> small models though right Mark when you think about god models versus small models as you were describing that but that would be small would you categorize that as a small >> well some of them I mean we I will let them announce you know them I will let them announce you know whatever they're doing whenever it's appropriate but some of them are now also doing big model development um and again this is also part of what this is also part of the learning just in the last two years well so like here's a big learning just from the last two years which is very interesting which is two years ago or three years ago for sure you would have said wow open AI is like way out ahead um and like it's probably going to be impossible for anybody to catch up and then it's like okay well Anthropic caught up and so but you know they came out of open AI and so they had all the secrets you know whatever and so knew how to do it and so okay they caught up but surely nobody can catch up after them and then very quickly after that there were a raft of other companies that caught up very fast and XAI is maybe the best example of that which is like you know XAI you know Elon's company XAI is the company name gro is the consumer product version of it um XAI basically caught up to you know state-of-the-art openai anthropic level in like less than 12 months from a standing start right and So, and again that kind of argues against any kind of permanent lead, right, by any one incumbent that's just going to basically be able to lock the entire market down like if you can catch up like that.","startTime":3281.68,"endTime":3291.119,"durationSeconds":9,"level":"C1","overallScore":9,"rationale":"후발 추격 가능성에 대한 핵심 통찰."},{"segmentIndex":13,"text":"I you know my view is we need to be actually very respectful of that and we need to be very aware of that and basically that we you know I use the metaphor with the dog that caught the bus like we always wanted to work on things that matter we are working on things that matter uh people in the rest of society actually really do care about these things um and you know and it's our responsibility to think that all through very carefully and to do a good job um you know both not just building the technology but also explaining it you know look you know I think we have a real obligation to uh you know to really explain ourselves and engage on these issues um in terms of how to measure how going you know it's sort of the classic social science question um uh which is like okay if you want to understand basically you know patterns of people there's basically two ways to understand what people are doing and thinking um one is to ask them and then the other is to watch them um and like every social scientist like every sociologist will tell you this which basically is you can ask people right and the way you do that right is like you know surveys focus groups polls um you know what they think Um but then you can watch them and you then but then you can watch them and you can do what's you know called reveal preferences.","startTime":4316.88,"endTime":4379.28,"durationSeconds":62,"level":"C2","overallScore":9,"rationale":"책임론과 사회과학 틀 제시."},{"segmentIndex":52,"text":"Um and you know you have these like long elaborate you know discussions about you know theories on this and that and the other thing and then reality just like completely smacks you square in the face you know like you idiot right you know like you know what were you like you know this is like the you know the ultimate frustration of the business which is also very motivating which is the number of times that you think that you've applied superior analysis and then you've either invested or not invested based on that analysis and it turns out it was just you the analysis was just completely wrong right um and you know you just like completely overrated your ability to epistemically you know kind of analyze these things you just you know basically inflicted harm like I always the question is always you know it's sort of you know any activity that we do is it value add or is it actually value subtract right and I think in this business of all businesses is kind of like that and that applies to all of my own contributions as well so there is that and then I would say um you know maybe the final thing is just like I do have the entire internet ready to tell me that I'm an idiot so that also doesn't hurt and it does on a regular basis on the point of uh your alluding to earlier about uh decisions on investing in companies.","startTime":4683.36,"endTime":4748.96,"durationSeconds":66,"level":"C2","overallScore":9.2,"rationale":"현실 검증과 오판의 교훈이 강함."},{"segmentIndex":2,"text":"We're seeing companies grow much faster. I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in 5 or 10 years. I think things are going to get much more sophisticated from here.","startTime":8.72,"endTime":16.96,"durationSeconds":8,"level":"C1","overallScore":7.8,"rationale":"제품 진화에 대한 전망이 뚜렷함."},{"segmentIndex":4,"text":"These are trillion dollar questions, not answers. But once somebody proves that it's capable, it seems to not be that hard for other people to be able to catch up, even people with far less resources.","startTime":20.8,"endTime":29.84,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"기술 추격의 본질을 짚는 문장."},{"segmentIndex":9,"text":"One is to ask them and then the other is to watch them. And what you often see in many areas of human activity, including politics and many different aspects of society, the answers that you get when you ask people are very different than the answers that you get when you watch them.","startTime":52.64,"endTime":64.159,"durationSeconds":12,"level":"C1","overallScore":8,"rationale":"말과 행동의 괴리를 통찰함."},{"segmentIndex":27,"text":"Um and so like and that was of course the path that the industry took which was building these kind of hyper literal you know mathematical machines you know that could execute mathematical operations billions of times per second but of course had no ability to kind of deal with human beings the way humans like to be dealt with and so you know couldn't understand you know human speech human language um and so forth and that's the computer industry that got built over the last 80 years and that's the computer industry that's built all the wealth of uh and financial returns of the computer industry uh over the last 80 years you know across all the generations of computers from mainframes through to smartphones.","startTime":174.64,"endTime":206.8,"durationSeconds":32,"level":"C2","overallScore":8,"rationale":"컴퓨터 산업의 한계를 깊게 설명."},{"segmentIndex":34,"text":"Um but and the neural network basically didn't happen but the neural network as an idea continued to be explored in academia um and sort of advanced research by sort of a rump you know movement that was originally called cybernetics and then became known as artificial intelligence uh basically for the last 80 years and essentially it didn't work like essentially it was basically decade after decade of excessive optimism uh followed by disappointment.","startTime":260.639,"endTime":283.199,"durationSeconds":23,"level":"C2","overallScore":8,"rationale":"AI 역사 사이클을 압축해 설명."},{"segmentIndex":41,"text":"So, we're sort of three year we're sort of three years in um to, you know, basically what is effectively an 80-year revolution um of actually being able to deliver on all the promise that the people on the all the on the alternate path, the sort of human cognition model path, you know, kind of saw from the very beginning and then, you know, the great news with this technology is it's already it's kind of ultra democratized.","startTime":325.36,"endTime":347.84000000000003,"durationSeconds":22,"level":"C2","overallScore":8,"rationale":"AI 혁명의 위치와 성격을 압축."},{"segmentIndex":47,"text":"Like I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in five or 10 years.","startTime":423.199,"endTime":426.56,"durationSeconds":3,"level":"C1","overallScore":7.8,"rationale":"미래 제품 변화 전망이 선명함."},{"segmentIndex":15,"text":"Um, and so sort of the internet's the carrier wave for AI to be able to proliferate at kind of light speed uh into the broad base of the global population.","startTime":682.88,"endTime":691.6,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"AI 확산 비유가 강하고 표현도 인상적."},{"segmentIndex":16,"text":"And that's a let's just say that's a potential rate of proliferation of a new technology that's just far faster than has ever been possible before.","startTime":691.6,"endTime":697.8389999999999,"durationSeconds":6,"level":"B2","overallScore":7.4,"rationale":"기술 확산 속도에 대한 통찰이 강함."},{"segmentIndex":19,"text":"Um, you know, you couldn't download television, but you can download AI. Um, and then and an incredible rate. Um, and then they're monetizing really well. Um and the core business model, right, is actually quite interesting. Um and then as a consequence there's kind of this hyperdelation of per unit cost and then that is like driving you know just like you know a more than corresponding level of demand growth you know with elasticity.","startTime":704.8,"endTime":715.2,"durationSeconds":10,"level":"C1","overallScore":8,"rationale":"사업모델·수요 논리가 압축돼 통찰 높음."},{"segmentIndex":25,"text":"Um, and you know, if you have the ability to like inject more intelligence into your business and you have the ability to do, you know, even the most prosaic things like raise your customer service scores, uh, you know, increase upsells, um, uh, you know, or reduce churn or if you have the ability to, um, you know, run marketing campaigns more effectively, um, you know, all of which AI is directly relevant to, like, you know, these are like direct business payoffs, um, you know, that people are seeing already.","startTime":764.48,"endTime":786.639,"durationSeconds":22,"level":"C1","overallScore":8.2,"rationale":"기업 효용 사례와 실전 표현이 매우 풍부함."}],"generatedAt":"2026-06-24T23:21:31.723Z","keyClipsTotalSec":2283},{"videoId":"xRh2sVcNXQ8","chunkIndex":7,"totalChunks":9,"title":"Marc Andreessen's 2026 Outlook: AI Timelines, US vs. China, and The Price of AI — Part 8 of 9","thumbnail":"https://i.ytimg.com/vi/xRh2sVcNXQ8/maxresdefault.jpg","duration":4878,"uploader":"a16z","youtubeUrl":"https://www.youtube.com/watch?v=xRh2sVcNXQ8","keywords":["ai","artificial-intelligence","technology-adoption","silicon-valley","venture-capital","public-opinion","automation","future-of-work","policy"],"normalizedKeywords":["비즈니스·전략","기술 트렌드","커리어·성장"],"targetAudience":[{"who":"창업자","why":"새 기술의 사회적 반발과 채택 간극을 이해하는 데 도움이 됨"},{"who":"투자자","why":"유행성 공포와 실제 사용 패턴의 차이를 읽는 관점이 유용함"},{"who":"지식노동자","why":"AI를 실제 업무에 어떻게 쓰는지와 대중 인식을 함께 볼 수 있음"}],"normalizedAudience":["창업자·스타트업","투자자·VC","지식노동자 일반"],"summary":"이 영상은 AI를 둘러싼 대중의 공포가 역사적으로 반복되어 온 과장된 반응이라고 보고, 실제 행동 데이터는 그와 정반대라는 점을 강조한다. 마르크스주의, 1960년대 AI 패닉, 2000년대 아웃소싱 공포 같은 사례를 들어 \"새 기술은 늘 일자리를 파괴한다\"는 서사가 매번 등장했지만 오래가지 않았다고 말한다.\n\n동시에 사람들은 설문에서는 AI를 두려워하면서도 실제로는 ChatGPT를 업무, 인간관계, 건강 문제 해결에 적극 활용하고 있다고 지적한다. 화자는 이런 \"말과 행동의 불일치\"가 당분간 공론장을 흔들겠지만, 결국 AI는 널리 퍼져 생활의 일부가 될 것이며, 기술을 만드는 쪽은 그것을 설명하고 사회적 맥락까지 책임져야 한다고 주장한다. 후반부에서는 투자 판단, 현실 감각, 젊은 세대에게 계속 배우는 태도, 그리고 VC가 겪는 반복적 후회까지 가볍게 이어진다.","insights":["새 기술에 대한 공포는 늘 반복되지만 오래가진 않는다.","설문상의 불안보다 실제 사용 데이터가 미래를 더 잘 말한다.","AI는 금지보다 확산이 먼저 오고, 뒤늦게 필수재가 된다.","기술을 만드는 쪽은 성능뿐 아니라 사회적 설명도 책임져야 한다.","투자와 판단은 결국 현실에 맞는지 빨리 검증되는 게임이다."],"keyClips":[{"clipId":"xRh2sVcNXQ8:c7:1-7","startSegmentIndex":1,"endSegmentIndex":7,"startTime":4200.55,"endTime":4289.6,"durationSeconds":89.1,"preview":"AI 패닉의 반복","mustSee":false},{"clipId":"xRh2sVcNXQ8:c7:8-13","startSegmentIndex":8,"endSegmentIndex":13,"startTime":4289.6,"endTime":4379.28,"durationSeconds":89.7,"preview":"기술의 사회적 책임","mustSee":false},{"clipId":"xRh2sVcNXQ8:c7:13-26","startSegmentIndex":13,"endSegmentIndex":26,"startTime":4316.88,"endTime":4520.719,"durationSeconds":203.8,"preview":"말과 행동의 괴리","mustSee":true},{"clipId":"xRh2sVcNXQ8:c7:42-52","startSegmentIndex":42,"endSegmentIndex":52,"startTime":4607.52,"endTime":4748.96,"durationSeconds":141.4,"preview":"현실 검증의 힘","mustSee":true},{"clipId":"xRh2sVcNXQ8:c7:53-60","startSegmentIndex":53,"endSegmentIndex":60,"startTime":4748.96,"endTime":4801.44,"durationSeconds":52.5,"preview":"VC의 후회 구조","mustSee":false}],"curatedSegments":[{"segmentIndex":45,"text":"Um, venture We have our issues and venture but a huge advantage that we have is we don't have to we can bet on multiple strategies at the same time right um and we are doing this so we are betting on big models and small models and prepared train models and open source models right and you know and foundation models and applications right uh and consumer and enterprise and so the portfolio approach the nature of it is like we are aggressively basically uh we are aggressively investing behind every strategy that we've identified that we think has a plausible chance of even when that even when that's contradictory to another strategy that we're investing in and one is just like the world's messy and probably a bunch of things are going to work and so like there's not going to be clean yes or no answers to a bunch of this like a lot a answers to a bunch of this like a lot of the answers to this I think are just going to be and answers but the other is like if one of these strategies doesn't work like you know we're not trying to hedge per se but you know we're going to have representation in the portfolio of the alternate strategy and so we're going to have mult multiple ways to win.","startTime":3451.52,"endTime":3506.48,"durationSeconds":55,"level":"C1","overallScore":10,"rationale":"벤처 포트폴리오 전략의 핵심 논리 제시."},{"segmentIndex":34,"text":"Um whereas of course China exports just a tremendous number of physical things right um including like a huge part of like the entire supply chain of parts that basically go into everything that American manufacturers you know kind of make right and so by the time a US you know whatever by the time an American company brings a toy to market right or a uh you know or a car um or anything or a computer or a smartphone or whatever like it's got a lot of componentry in it that was made in China so there is a much tighter in interlinkage between the American and Chinese economies than there as the American and Soviet economies and you know may maybe you know Adam Smith or whatever might say you know that's good news for peace and that you know both countries need each other by the way the other part of that argument is that the Chinese basically the Chinese you know the Chinese governance model is based on high employment um you know because you know if you know at least all the geopolitical people say if China ended up with like 25 or 50% unemployment that would cause civil unrest which is the one thing that the CCP doesn't want and so the corresponding part of the trade pressure is China needs the American export market you know the American consumer is like a third of the global economy.","startTime":1535.52,"endTime":1595.84,"durationSeconds":60,"level":"C2","overallScore":8.8,"rationale":"무역 상호의존과 정치 논리를 깊게 설명."},{"segmentIndex":5,"text":"it came from a hedge fund and it like as far as we can tell it basically is this somewhat idiosyncratic situation where you just have this incredibly successful quant hedge fund with all these you know super geniuses um and the founder of that hedge fund you know basically decided to build AI um and you know at least external indications are this was a surprise to even the Chinese government it's impossible to prove you know what the Chinese government was surprised by or not but you know there's at least the atmospherics are that this was not exactly planned this was not a national champion tech company at the time that Deepseek was released it was it sort of came out of left field which by the way is very encouraging for the field that it was possible for somebody to do that kind of who was unknown right because it kind of means that maybe you don't need all these you know super genius superstar researchers maybe actually smart kids can just build this stuff which I think is the direction things are headed um and so that kicked off I would say like this kind of I don't know copycat's the wrong word but that was sort of it feels like the success of deepseek and the success of deepseek from China as open source kind of kicked off a sort of trend in China releasing these open source models um you know Look, the cynics, you know, in DC would say, you know, yeah, like they're dumping, right?","startTime":1839.12,"endTime":1900.72,"durationSeconds":62,"level":"C1","overallScore":9,"rationale":"배경 해석과 관점, 표현이 매우 풍부."},{"segmentIndex":60,"text":"um you know which is a key way that all this technology proliferates and so this law would have assigned downstream liability to any misuse of open source to the original developer of the open source and so you know you're an independent developer or you're an academic or you're a startup you develop and release an AI model the AI model works fine the day you release it it's great but like 5 years later it gets built into a nuclear power plant and then there's a meltdown of the nuclear","startTime":2379.76,"endTime":2389.76,"durationSeconds":10,"level":"C1","overallScore":9,"rationale":"오픈소스 책임 전가의 문제를 선명히 설명."},{"segmentIndex":27,"text":"Um, and so you've got like the most magic new technology in the world and then it's basically being served up by those companies in a as a cloud business and made basically available to everybody on the planet to just click and use and for like relatively small amounts of money and then on a usage basis which means and usage is great for startups because you it means you can start easily right you the you know there's very you know there's basically no fixed co for a startup building an AI app they don't have giant fixed cost because they could just tap into the open AI or anthropic or Google or Microsoft or whatever you know cloud you know tokens by the you know, intelligence tokens by the drink offering and just get going.","startTime":2607.44,"endTime":2639.839,"durationSeconds":32,"level":"C2","overallScore":9,"rationale":"스타트업 비용구조 통찰과 표현이 풍부."},{"segmentIndex":22,"text":"they're actually they actually as these products get more sophisticated they actually end up using many different kinds of models that are kind of customtailored to the specific aspects of how these products work. Um and so they may start out using one model but they end up using a dozen models and then in the fullness of time it might be 50 or 100 different models for different aspects of the product. Um and so they a lot of these the leading edge application companies are actually backward integrating and actually building their own AI models because they have the deepest understanding of their domain.","startTime":3222.48,"endTime":3234.72,"durationSeconds":12,"level":"C1","overallScore":9,"rationale":"앱 기업의 진화 논리를 깊게 설명."},{"segmentIndex":27,"text":"Um and so that you know that's I think >> small models though right Mark when you think about god models versus small models as you were describing that but that would be small would you categorize that as a small >> well some of them I mean we I will let them announce you know them I will let them announce you know whatever they're doing whenever it's appropriate but some of them are now also doing big model development um and again this is also part of what this is also part of the learning just in the last two years well so like here's a big learning just from the last two years which is very interesting which is two years ago or three years ago for sure you would have said wow open AI is like way out ahead um and like it's probably going to be impossible for anybody to catch up and then it's like okay well Anthropic caught up and so but you know they came out of open AI and so they had all the secrets you know whatever and so knew how to do it and so okay they caught up but surely nobody can catch up after them and then very quickly after that there were a raft of other companies that caught up very fast and XAI is maybe the best example of that which is like you know XAI you know Elon's company XAI is the company name gro is the consumer product version of it um XAI basically caught up to you know state-of-the-art openai anthropic level in like less than 12 months from a standing start right and So, and again that kind of argues against any kind of permanent lead, right, by any one incumbent that's just going to basically be able to lock the entire market down like if you can catch up like that.","startTime":3281.68,"endTime":3291.119,"durationSeconds":9,"level":"C1","overallScore":9,"rationale":"후발 추격 가능성에 대한 핵심 통찰."},{"segmentIndex":13,"text":"I you know my view is we need to be actually very respectful of that and we need to be very aware of that and basically that we you know I use the metaphor with the dog that caught the bus like we always wanted to work on things that matter we are working on things that matter uh people in the rest of society actually really do care about these things um and you know and it's our responsibility to think that all through very carefully and to do a good job um you know both not just building the technology but also explaining it you know look you know I think we have a real obligation to uh you know to really explain ourselves and engage on these issues um in terms of how to measure how going you know it's sort of the classic social science question um uh which is like okay if you want to understand basically you know patterns of people there's basically two ways to understand what people are doing and thinking um one is to ask them and then the other is to watch them um and like every social scientist like every sociologist will tell you this which basically is you can ask people right and the way you do that right is like you know surveys focus groups polls um you know what they think Um but then you can watch them and you then but then you can watch them and you can do what's you know called reveal preferences.","startTime":4316.88,"endTime":4379.28,"durationSeconds":62,"level":"C2","overallScore":9,"rationale":"책임론과 사회과학 틀 제시."},{"segmentIndex":52,"text":"Um and you know you have these like long elaborate you know discussions about you know theories on this and that and the other thing and then reality just like completely smacks you square in the face you know like you idiot right you know like you know what were you like you know this is like the you know the ultimate frustration of the business which is also very motivating which is the number of times that you think that you've applied superior analysis and then you've either invested or not invested based on that analysis and it turns out it was just you the analysis was just completely wrong right um and you know you just like completely overrated your ability to epistemically you know kind of analyze these things you just you know basically inflicted harm like I always the question is always you know it's sort of you know any activity that we do is it value add or is it actually value subtract right and I think in this business of all businesses is kind of like that and that applies to all of my own contributions as well so there is that and then I would say um you know maybe the final thing is just like I do have the entire internet ready to tell me that I'm an idiot so that also doesn't hurt and it does on a regular basis on the point of uh your alluding to earlier about uh decisions on investing in companies.","startTime":4683.36,"endTime":4748.96,"durationSeconds":66,"level":"C2","overallScore":9.2,"rationale":"현실 검증과 오판의 교훈이 강함."},{"segmentIndex":2,"text":"We're seeing companies grow much faster. I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in 5 or 10 years. I think things are going to get much more sophisticated from here.","startTime":8.72,"endTime":16.96,"durationSeconds":8,"level":"C1","overallScore":7.8,"rationale":"제품 진화에 대한 전망이 뚜렷함."},{"segmentIndex":4,"text":"These are trillion dollar questions, not answers. But once somebody proves that it's capable, it seems to not be that hard for other people to be able to catch up, even people with far less resources.","startTime":20.8,"endTime":29.84,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"기술 추격의 본질을 짚는 문장."},{"segmentIndex":9,"text":"One is to ask them and then the other is to watch them. And what you often see in many areas of human activity, including politics and many different aspects of society, the answers that you get when you ask people are very different than the answers that you get when you watch them.","startTime":52.64,"endTime":64.159,"durationSeconds":12,"level":"C1","overallScore":8,"rationale":"말과 행동의 괴리를 통찰함."},{"segmentIndex":27,"text":"Um and so like and that was of course the path that the industry took which was building these kind of hyper literal you know mathematical machines you know that could execute mathematical operations billions of times per second but of course had no ability to kind of deal with human beings the way humans like to be dealt with and so you know couldn't understand you know human speech human language um and so forth and that's the computer industry that got built over the last 80 years and that's the computer industry that's built all the wealth of uh and financial returns of the computer industry uh over the last 80 years you know across all the generations of computers from mainframes through to smartphones.","startTime":174.64,"endTime":206.8,"durationSeconds":32,"level":"C2","overallScore":8,"rationale":"컴퓨터 산업의 한계를 깊게 설명."},{"segmentIndex":34,"text":"Um but and the neural network basically didn't happen but the neural network as an idea continued to be explored in academia um and sort of advanced research by sort of a rump you know movement that was originally called cybernetics and then became known as artificial intelligence uh basically for the last 80 years and essentially it didn't work like essentially it was basically decade after decade of excessive optimism uh followed by disappointment.","startTime":260.639,"endTime":283.199,"durationSeconds":23,"level":"C2","overallScore":8,"rationale":"AI 역사 사이클을 압축해 설명."},{"segmentIndex":41,"text":"So, we're sort of three year we're sort of three years in um to, you know, basically what is effectively an 80-year revolution um of actually being able to deliver on all the promise that the people on the all the on the alternate path, the sort of human cognition model path, you know, kind of saw from the very beginning and then, you know, the great news with this technology is it's already it's kind of ultra democratized.","startTime":325.36,"endTime":347.84000000000003,"durationSeconds":22,"level":"C2","overallScore":8,"rationale":"AI 혁명의 위치와 성격을 압축."},{"segmentIndex":47,"text":"Like I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in five or 10 years.","startTime":423.199,"endTime":426.56,"durationSeconds":3,"level":"C1","overallScore":7.8,"rationale":"미래 제품 변화 전망이 선명함."},{"segmentIndex":15,"text":"Um, and so sort of the internet's the carrier wave for AI to be able to proliferate at kind of light speed uh into the broad base of the global population.","startTime":682.88,"endTime":691.6,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"AI 확산 비유가 강하고 표현도 인상적."},{"segmentIndex":16,"text":"And that's a let's just say that's a potential rate of proliferation of a new technology that's just far faster than has ever been possible before.","startTime":691.6,"endTime":697.8389999999999,"durationSeconds":6,"level":"B2","overallScore":7.4,"rationale":"기술 확산 속도에 대한 통찰이 강함."},{"segmentIndex":19,"text":"Um, you know, you couldn't download television, but you can download AI. Um, and then and an incredible rate. Um, and then they're monetizing really well. Um and the core business model, right, is actually quite interesting. Um and then as a consequence there's kind of this hyperdelation of per unit cost and then that is like driving you know just like you know a more than corresponding level of demand growth you know with elasticity.","startTime":704.8,"endTime":715.2,"durationSeconds":10,"level":"C1","overallScore":8,"rationale":"사업모델·수요 논리가 압축돼 통찰 높음."},{"segmentIndex":25,"text":"Um, and you know, if you have the ability to like inject more intelligence into your business and you have the ability to do, you know, even the most prosaic things like raise your customer service scores, uh, you know, increase upsells, um, uh, you know, or reduce churn or if you have the ability to, um, you know, run marketing campaigns more effectively, um, you know, all of which AI is directly relevant to, like, you know, these are like direct business payoffs, um, you know, that people are seeing already.","startTime":764.48,"endTime":786.639,"durationSeconds":22,"level":"C1","overallScore":8.2,"rationale":"기업 효용 사례와 실전 표현이 매우 풍부함."}],"generatedAt":"2026-06-24T23:21:52.703Z","keyClipsTotalSec":2283},{"videoId":"xRh2sVcNXQ8","chunkIndex":8,"totalChunks":9,"title":"Marc Andreessen's 2026 Outlook: AI Timelines, US vs. China, and The Price of AI — Part 9 of 9","thumbnail":"https://i.ytimg.com/vi/xRh2sVcNXQ8/maxresdefault.jpg","duration":4878,"uploader":"a16z","youtubeUrl":"https://www.youtube.com/watch?v=xRh2sVcNXQ8","keywords":["mars","spaceflight","elon-musk","futurism","ai","technology","innovation","future-society","venture-capital"],"normalizedKeywords":["기술 트렌드","비즈니스·전략"],"targetAudience":[{"who":"예비 창업자","why":"기술이 현실을 바꾸는 속도와 확률을 감각적으로 이해할 수 있음"},{"who":"투자 관심자","why":"장기 기술 테마의 타임라인과 확률적 전망을 읽는 데 도움됨"},{"who":"호기심 많은 직장인","why":"AI와 우주기술의 미래를 가볍게 따라가며 시야를 넓힐 수 있음"}],"normalizedAudience":["창업자·스타트업","투자자·VC","지식노동자 일반"],"summary":"이 영상의 말미 구간은 인터뷰어가 마르스 이주 의사를 묻고, 화자가 개인적으로는 가기 싫지만 기술적으로는 머지않아 왕복 여행이 일상화될 수 있다고 보는 대화를 담고 있다. 화자는 엘론 머스크가 결국 해낼 것이라고 보면서도, 정확한 시점을 단정하지는 않으려는 태도를 유지한다.\n\n핵심은 미래 기술을 낙관하되, 과장된 확신 대신 확률적·시간축 관점으로 보는 것이다. 머지않아 우주 여행이 '이론적 상상'이 아니라 실제 선택지가 될 수 있고, 그러면 개인의 선호조차 현실적인 의사결정 문제가 된다는 점을 암시한다.","insights":["기술 예측은 단정보다 확률이 더 정확하다.","불가능해 보이는 것도 시간축이 바뀌면 현실이 된다.","개인의 호불호도 기술이 성숙하면 실무 문제가 된다.","장기 혁신은 창업자의 집요함이 성패를 가른다."],"keyClips":[{"clipId":"xRh2sVcNXQ8:c8:2-6","startSegmentIndex":2,"endSegmentIndex":6,"startTime":4805.92,"endTime":4832.8,"durationSeconds":26.9,"preview":"마르스는 아직 먼가","mustSee":false},{"clipId":"xRh2sVcNXQ8:c8:7-10","startSegmentIndex":7,"endSegmentIndex":10,"startTime":4832.8,"endTime":4850.08,"durationSeconds":17.3,"preview":"예측보다 가능성","mustSee":false},{"clipId":"xRh2sVcNXQ8:c8:11-12","startSegmentIndex":11,"endSegmentIndex":12,"startTime":4850.08,"endTime":4867.12,"durationSeconds":17,"preview":"우주여행의 일상화","mustSee":false}],"curatedSegments":[{"segmentIndex":45,"text":"Um, venture We have our issues and venture but a huge advantage that we have is we don't have to we can bet on multiple strategies at the same time right um and we are doing this so we are betting on big models and small models and prepared train models and open source models right and you know and foundation models and applications right uh and consumer and enterprise and so the portfolio approach the nature of it is like we are aggressively basically uh we are aggressively investing behind every strategy that we've identified that we think has a plausible chance of even when that even when that's contradictory to another strategy that we're investing in and one is just like the world's messy and probably a bunch of things are going to work and so like there's not going to be clean yes or no answers to a bunch of this like a lot a answers to a bunch of this like a lot of the answers to this I think are just going to be and answers but the other is like if one of these strategies doesn't work like you know we're not trying to hedge per se but you know we're going to have representation in the portfolio of the alternate strategy and so we're going to have mult multiple ways to win.","startTime":3451.52,"endTime":3506.48,"durationSeconds":55,"level":"C1","overallScore":10,"rationale":"벤처 포트폴리오 전략의 핵심 논리 제시."},{"segmentIndex":34,"text":"Um whereas of course China exports just a tremendous number of physical things right um including like a huge part of like the entire supply chain of parts that basically go into everything that American manufacturers you know kind of make right and so by the time a US you know whatever by the time an American company brings a toy to market right or a uh you know or a car um or anything or a computer or a smartphone or whatever like it's got a lot of componentry in it that was made in China so there is a much tighter in interlinkage between the American and Chinese economies than there as the American and Soviet economies and you know may maybe you know Adam Smith or whatever might say you know that's good news for peace and that you know both countries need each other by the way the other part of that argument is that the Chinese basically the Chinese you know the Chinese governance model is based on high employment um you know because you know if you know at least all the geopolitical people say if China ended up with like 25 or 50% unemployment that would cause civil unrest which is the one thing that the CCP doesn't want and so the corresponding part of the trade pressure is China needs the American export market you know the American consumer is like a third of the global economy.","startTime":1535.52,"endTime":1595.84,"durationSeconds":60,"level":"C2","overallScore":8.8,"rationale":"무역 상호의존과 정치 논리를 깊게 설명."},{"segmentIndex":5,"text":"it came from a hedge fund and it like as far as we can tell it basically is this somewhat idiosyncratic situation where you just have this incredibly successful quant hedge fund with all these you know super geniuses um and the founder of that hedge fund you know basically decided to build AI um and you know at least external indications are this was a surprise to even the Chinese government it's impossible to prove you know what the Chinese government was surprised by or not but you know there's at least the atmospherics are that this was not exactly planned this was not a national champion tech company at the time that Deepseek was released it was it sort of came out of left field which by the way is very encouraging for the field that it was possible for somebody to do that kind of who was unknown right because it kind of means that maybe you don't need all these you know super genius superstar researchers maybe actually smart kids can just build this stuff which I think is the direction things are headed um and so that kicked off I would say like this kind of I don't know copycat's the wrong word but that was sort of it feels like the success of deepseek and the success of deepseek from China as open source kind of kicked off a sort of trend in China releasing these open source models um you know Look, the cynics, you know, in DC would say, you know, yeah, like they're dumping, right?","startTime":1839.12,"endTime":1900.72,"durationSeconds":62,"level":"C1","overallScore":9,"rationale":"배경 해석과 관점, 표현이 매우 풍부."},{"segmentIndex":60,"text":"um you know which is a key way that all this technology proliferates and so this law would have assigned downstream liability to any misuse of open source to the original developer of the open source and so you know you're an independent developer or you're an academic or you're a startup you develop and release an AI model the AI model works fine the day you release it it's great but like 5 years later it gets built into a nuclear power plant and then there's a meltdown of the nuclear","startTime":2379.76,"endTime":2389.76,"durationSeconds":10,"level":"C1","overallScore":9,"rationale":"오픈소스 책임 전가의 문제를 선명히 설명."},{"segmentIndex":27,"text":"Um, and so you've got like the most magic new technology in the world and then it's basically being served up by those companies in a as a cloud business and made basically available to everybody on the planet to just click and use and for like relatively small amounts of money and then on a usage basis which means and usage is great for startups because you it means you can start easily right you the you know there's very you know there's basically no fixed co for a startup building an AI app they don't have giant fixed cost because they could just tap into the open AI or anthropic or Google or Microsoft or whatever you know cloud you know tokens by the you know, intelligence tokens by the drink offering and just get going.","startTime":2607.44,"endTime":2639.839,"durationSeconds":32,"level":"C2","overallScore":9,"rationale":"스타트업 비용구조 통찰과 표현이 풍부."},{"segmentIndex":22,"text":"they're actually they actually as these products get more sophisticated they actually end up using many different kinds of models that are kind of customtailored to the specific aspects of how these products work. Um and so they may start out using one model but they end up using a dozen models and then in the fullness of time it might be 50 or 100 different models for different aspects of the product. Um and so they a lot of these the leading edge application companies are actually backward integrating and actually building their own AI models because they have the deepest understanding of their domain.","startTime":3222.48,"endTime":3234.72,"durationSeconds":12,"level":"C1","overallScore":9,"rationale":"앱 기업의 진화 논리를 깊게 설명."},{"segmentIndex":27,"text":"Um and so that you know that's I think >> small models though right Mark when you think about god models versus small models as you were describing that but that would be small would you categorize that as a small >> well some of them I mean we I will let them announce you know them I will let them announce you know whatever they're doing whenever it's appropriate but some of them are now also doing big model development um and again this is also part of what this is also part of the learning just in the last two years well so like here's a big learning just from the last two years which is very interesting which is two years ago or three years ago for sure you would have said wow open AI is like way out ahead um and like it's probably going to be impossible for anybody to catch up and then it's like okay well Anthropic caught up and so but you know they came out of open AI and so they had all the secrets you know whatever and so knew how to do it and so okay they caught up but surely nobody can catch up after them and then very quickly after that there were a raft of other companies that caught up very fast and XAI is maybe the best example of that which is like you know XAI you know Elon's company XAI is the company name gro is the consumer product version of it um XAI basically caught up to you know state-of-the-art openai anthropic level in like less than 12 months from a standing start right and So, and again that kind of argues against any kind of permanent lead, right, by any one incumbent that's just going to basically be able to lock the entire market down like if you can catch up like that.","startTime":3281.68,"endTime":3291.119,"durationSeconds":9,"level":"C1","overallScore":9,"rationale":"후발 추격 가능성에 대한 핵심 통찰."},{"segmentIndex":13,"text":"I you know my view is we need to be actually very respectful of that and we need to be very aware of that and basically that we you know I use the metaphor with the dog that caught the bus like we always wanted to work on things that matter we are working on things that matter uh people in the rest of society actually really do care about these things um and you know and it's our responsibility to think that all through very carefully and to do a good job um you know both not just building the technology but also explaining it you know look you know I think we have a real obligation to uh you know to really explain ourselves and engage on these issues um in terms of how to measure how going you know it's sort of the classic social science question um uh which is like okay if you want to understand basically you know patterns of people there's basically two ways to understand what people are doing and thinking um one is to ask them and then the other is to watch them um and like every social scientist like every sociologist will tell you this which basically is you can ask people right and the way you do that right is like you know surveys focus groups polls um you know what they think Um but then you can watch them and you then but then you can watch them and you can do what's you know called reveal preferences.","startTime":4316.88,"endTime":4379.28,"durationSeconds":62,"level":"C2","overallScore":9,"rationale":"책임론과 사회과학 틀 제시."},{"segmentIndex":52,"text":"Um and you know you have these like long elaborate you know discussions about you know theories on this and that and the other thing and then reality just like completely smacks you square in the face you know like you idiot right you know like you know what were you like you know this is like the you know the ultimate frustration of the business which is also very motivating which is the number of times that you think that you've applied superior analysis and then you've either invested or not invested based on that analysis and it turns out it was just you the analysis was just completely wrong right um and you know you just like completely overrated your ability to epistemically you know kind of analyze these things you just you know basically inflicted harm like I always the question is always you know it's sort of you know any activity that we do is it value add or is it actually value subtract right and I think in this business of all businesses is kind of like that and that applies to all of my own contributions as well so there is that and then I would say um you know maybe the final thing is just like I do have the entire internet ready to tell me that I'm an idiot so that also doesn't hurt and it does on a regular basis on the point of uh your alluding to earlier about uh decisions on investing in companies.","startTime":4683.36,"endTime":4748.96,"durationSeconds":66,"level":"C2","overallScore":9.2,"rationale":"현실 검증과 오판의 교훈이 강함."},{"segmentIndex":2,"text":"We're seeing companies grow much faster. I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in 5 or 10 years. I think things are going to get much more sophisticated from here.","startTime":8.72,"endTime":16.96,"durationSeconds":8,"level":"C1","overallScore":7.8,"rationale":"제품 진화에 대한 전망이 뚜렷함."},{"segmentIndex":4,"text":"These are trillion dollar questions, not answers. But once somebody proves that it's capable, it seems to not be that hard for other people to be able to catch up, even people with far less resources.","startTime":20.8,"endTime":29.84,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"기술 추격의 본질을 짚는 문장."},{"segmentIndex":9,"text":"One is to ask them and then the other is to watch them. And what you often see in many areas of human activity, including politics and many different aspects of society, the answers that you get when you ask people are very different than the answers that you get when you watch them.","startTime":52.64,"endTime":64.159,"durationSeconds":12,"level":"C1","overallScore":8,"rationale":"말과 행동의 괴리를 통찰함."},{"segmentIndex":27,"text":"Um and so like and that was of course the path that the industry took which was building these kind of hyper literal you know mathematical machines you know that could execute mathematical operations billions of times per second but of course had no ability to kind of deal with human beings the way humans like to be dealt with and so you know couldn't understand you know human speech human language um and so forth and that's the computer industry that got built over the last 80 years and that's the computer industry that's built all the wealth of uh and financial returns of the computer industry uh over the last 80 years you know across all the generations of computers from mainframes through to smartphones.","startTime":174.64,"endTime":206.8,"durationSeconds":32,"level":"C2","overallScore":8,"rationale":"컴퓨터 산업의 한계를 깊게 설명."},{"segmentIndex":34,"text":"Um but and the neural network basically didn't happen but the neural network as an idea continued to be explored in academia um and sort of advanced research by sort of a rump you know movement that was originally called cybernetics and then became known as artificial intelligence uh basically for the last 80 years and essentially it didn't work like essentially it was basically decade after decade of excessive optimism uh followed by disappointment.","startTime":260.639,"endTime":283.199,"durationSeconds":23,"level":"C2","overallScore":8,"rationale":"AI 역사 사이클을 압축해 설명."},{"segmentIndex":41,"text":"So, we're sort of three year we're sort of three years in um to, you know, basically what is effectively an 80-year revolution um of actually being able to deliver on all the promise that the people on the all the on the alternate path, the sort of human cognition model path, you know, kind of saw from the very beginning and then, you know, the great news with this technology is it's already it's kind of ultra democratized.","startTime":325.36,"endTime":347.84000000000003,"durationSeconds":22,"level":"C2","overallScore":8,"rationale":"AI 혁명의 위치와 성격을 압축."},{"segmentIndex":47,"text":"Like I'm very skeptical that the form and shape of the products that people are using today is what they're going to be using in five or 10 years.","startTime":423.199,"endTime":426.56,"durationSeconds":3,"level":"C1","overallScore":7.8,"rationale":"미래 제품 변화 전망이 선명함."},{"segmentIndex":15,"text":"Um, and so sort of the internet's the carrier wave for AI to be able to proliferate at kind of light speed uh into the broad base of the global population.","startTime":682.88,"endTime":691.6,"durationSeconds":9,"level":"C1","overallScore":7.8,"rationale":"AI 확산 비유가 강하고 표현도 인상적."},{"segmentIndex":16,"text":"And that's a let's just say that's a potential rate of proliferation of a new technology that's just far faster than has ever been possible before.","startTime":691.6,"endTime":697.8389999999999,"durationSeconds":6,"level":"B2","overallScore":7.4,"rationale":"기술 확산 속도에 대한 통찰이 강함."},{"segmentIndex":19,"text":"Um, you know, you couldn't download television, but you can download AI. Um, and then and an incredible rate. Um, and then they're monetizing really well. Um and the core business model, right, is actually quite interesting. Um and then as a consequence there's kind of this hyperdelation of per unit cost and then that is like driving you know just like you know a more than corresponding level of demand growth you know with elasticity.","startTime":704.8,"endTime":715.2,"durationSeconds":10,"level":"C1","overallScore":8,"rationale":"사업모델·수요 논리가 압축돼 통찰 높음."},{"segmentIndex":25,"text":"Um, and you know, if you have the ability to like inject more intelligence into your business and you have the ability to do, you know, even the most prosaic things like raise your customer service scores, uh, you know, increase upsells, um, uh, you know, or reduce churn or if you have the ability to, um, you know, run marketing campaigns more effectively, um, you know, all of which AI is directly relevant to, like, you know, these are like direct business payoffs, um, you know, that people are seeing already.","startTime":764.48,"endTime":786.639,"durationSeconds":22,"level":"C1","overallScore":8.2,"rationale":"기업 효용 사례와 실전 표현이 매우 풍부함."}],"generatedAt":"2026-06-24T23:22:22.531Z","keyClipsTotalSec":2283}]}