{"success":true,"count":3,"items":[{"videoId":"7zCsfe57tpU","chunkIndex":0,"totalChunks":3,"title":"Is this the only skill left? — Part 1 of 3","thumbnail":"https://i.ytimg.com/vi/7zCsfe57tpU/maxresdefault.jpg","duration":1330,"uploader":"Hak","youtubeUrl":"https://www.youtube.com/watch?v=7zCsfe57tpU","keywords":["systems-thinking","software-engineering","ai-coding","developer-skills","architecture","agentic-ai","cognitive-debt","abstraction","product-development","technical-leadership"],"normalizedKeywords":["엔지니어링","커리어·성장","기술 트렌드"],"targetAudience":[{"who":"엔지니어","why":"AI가 코드 생성하는 시대에 시스템을 설계·검증하는 기준이 필요함"},{"who":"주니어 개발자","why":"코딩보다 중요한 판단·리뷰·블라스트 반경 감각을 배울 수 있음"},{"who":"비기술 창업자","why":"AI로 만든 제품의 구조적 위험과 운영 실패를 이해하는 데 유용함"}],"normalizedAudience":["엔지니어·개발자","학생·주니어","창업자·스타트업"],"summary":"이 영상은 AI가 코드를 쉽게 만들어주는 시대일수록, 개발자에게 가장 중요한 능력은 '프롬프트 작성'이 아니라 시스템을 이해하고 설계하는 능력이라고 주장한다. 피터 나우르의 '코드는 그림자일 뿐, 진짜 프로그램은 개발자의 머릿속에 있다'는 관점을 가져와, AI가 코드를 대신 써도 시스템의 연결 구조와 책임, 피드백, 삭제 시 영향 범위는 사람이 이해해야 한다고 강조한다.\n\n또한 LLM은 컴파일러처럼 검증 가능한 추상화가 아니라 확률적 협력자이므로, 결과를 맹신하면 이해 부채와 시스템 붕괴가 생긴다고 말한다. 실제로 AI로 만든 앱을 감사한 사례를 통해, 로그 부재·상태 분산·에러 처리 부재 같은 문제는 코딩 실력보다 시스템 사고 부족에서 비롯된다고 보여준다. 결론적으로, AI 시대의 개발자는 '코드를 잘 쓰는 사람'보다 '전체가 어떻게 작동하는지 판단하는 사람'이어야 한다는 메시지다.","insights":["AI 시대엔 코딩보다 시스템을 보는 눈이 먼저다.","좋은 프롬프트보다 상태·피드백·의존성 파악이 더 중요하다.","LLM은 검증된 추상화가 아니라 이해해야 믿을 수 있는 도구다.","많은 장애는 버그보다 설계 부재에서 먼저 터진다.","주니어도 첫날부터 블라스트 반경을 생각해야 한다."],"keyClips":[{"clipId":"7zCsfe57tpU:c0:1-18","startSegmentIndex":1,"endSegmentIndex":18,"startTime":0.30000000000000004,"endTime":113.52052631578947,"durationSeconds":113.2,"preview":"코드보다 시스템","mustSee":true},{"clipId":"7zCsfe57tpU:c0:19-40","startSegmentIndex":19,"endSegmentIndex":40,"startTime":113.42,"endTime":262.47105263157897,"durationSeconds":149.1,"preview":"운전석은 사람","mustSee":false},{"clipId":"7zCsfe57tpU:c0:41-57","startSegmentIndex":41,"endSegmentIndex":57,"startTime":262.14,"endTime":376.7535294117647,"durationSeconds":114.6,"preview":"새로운 개발 문법","mustSee":true},{"clipId":"7zCsfe57tpU:c0:58-70","startSegmentIndex":58,"endSegmentIndex":70,"startTime":376.62,"endTime":460.6291176470588,"durationSeconds":84,"preview":"실패한 AI 앱","mustSee":false},{"clipId":"7zCsfe57tpU:c0:71-92","startSegmentIndex":71,"endSegmentIndex":92,"startTime":460.53999999999996,"endTime":603.9705882352941,"durationSeconds":143.4,"preview":"AI는 추상화 아님","mustSee":true}],"curatedSegments":[{"segmentIndex":13,"text":"program. The code is just the shadow of it, as Peter Naur puts it. Fourty years later, today we","startTime":73.26,"endTime":80.27,"durationSeconds":7,"level":"B2","overallScore":7.6,"rationale":"비유가 강하고 핵심 메시지도 뚜렷."},{"segmentIndex":20,"text":"career asking \"if AI writes the code, how do we build the judgment to evaluate it? Should I","startTime":119.18,"endTime":125.79,"durationSeconds":7,"level":"B2","overallScore":7.8,"rationale":"학습자 고민을 직접 던지는 핵심 질문."},{"segmentIndex":38,"text":"how the parts fit together, when the strings should hold back, when the brass should come in.","startTime":241.26000000000002,"endTime":247.14794117647057,"durationSeconds":6,"level":"B2","overallScore":7.8,"rationale":"조율의 본질을 구체적으로 보여줌."},{"segmentIndex":55,"text":"If nothing tells you that, the system isn't working. Probably is pretending to work. Three,","startTime":356.38,"endTime":362.58,"durationSeconds":6,"level":"B2","overallScore":7.6,"rationale":"관측 없는 시스템의 위험을 날카롭게 짚음."},{"segmentIndex":57,"text":"the system in your head before you touch it? That's the theory Naur was talking about. Now,","startTime":369.26,"endTime":376.7535294117647,"durationSeconds":7,"level":"B2","overallScore":7.6,"rationale":"이론의 실천적 의미를 연결해 줌."},{"segmentIndex":65,"text":"there was a systems thinking failure, not a coding one. Where does state live? Well, everywhere all","startTime":423.34,"endTime":429.5938235294118,"durationSeconds":6,"level":"C1","overallScore":7.8,"rationale":""},{"segmentIndex":79,"text":"just a probabilistic translator. It's stochastic by nature. For the same input, you get different","startTime":513.18,"endTime":520.0236666666667,"durationSeconds":7,"level":"C1","overallScore":7.8,"rationale":"LLM의 본질을 대비적으로 설명한다."},{"segmentIndex":80,"text":"outputs almost each time. No guarantees. This compiler could introduce a security vulnerability,","startTime":519.8199999999999,"endTime":526.6507692307692,"durationSeconds":7,"level":"C1","overallScore":7.6,"rationale":"구체적 위험 예시로 설득력이 높다."},{"segmentIndex":17,"text":"now they need them back. So if you're a junior, you're not behind. You just got robbed of the","startTime":712.22,"endTime":717.2465789473686,"durationSeconds":5,"level":"B1","overallScore":6.8,"rationale":""},{"segmentIndex":26,"text":"one. And I think yes, there will be fewer, but the ones who will make it will be more valuable than","startTime":773.1,"endTime":782.3028571428572,"durationSeconds":9,"level":"B1","overallScore":6.8,"rationale":""},{"segmentIndex":59,"text":"great learning partner if you use it correctly. Yes, prompt, generate, but then study what comes","startTime":992.22,"endTime":998.7450000000001,"durationSeconds":7,"level":"B2","overallScore":7.8,"rationale":"AI 학습 활용법을 구체적으로 제안함."},{"segmentIndex":60,"text":"back. Treat the AI like an infinitely patient senior dev and ask it \"why this approach? What","startTime":998.62,"endTime":1005.5976470588237,"durationSeconds":7,"level":"C1","overallScore":8,"rationale":"질문 방식까지 제시한 고가치 조언."},{"segmentIndex":64,"text":"The next generation of seniors will be juniors who use AI to understand better and faster,","startTime":1022.9399999999999,"endTime":1028.325,"durationSeconds":5,"level":"B2","overallScore":7.4,"rationale":""},{"segmentIndex":11,"text":"the program. The program is what lives inside the programmer's head. How the pieces connect,","startTime":61.98,"endTime":67.68866666666666,"durationSeconds":6,"level":"B2","overallScore":6,"rationale":""},{"segmentIndex":64,"text":"error handling, but the app was apparently working until it didn't. Every single failure mode in","startTime":416.14,"endTime":423.39,"durationSeconds":7,"level":"B2","overallScore":6.6,"rationale":"시스템 취약성을 인상적으로 드러낸다."},{"segmentIndex":66,"text":"at once. Where does feedback live? Nowhere. We're blind. What breaks if I delete this? Never asked,","startTime":429.5,"endTime":436.73294117647055,"durationSeconds":7,"level":"B2","overallScore":7.2,"rationale":""},{"segmentIndex":78,"text":"it did the right thing. That's what makes it a trustworthy abstraction. An LLM is not that. It's","startTime":505.65999999999997,"endTime":513.2711111111112,"durationSeconds":8,"level":"C1","overallScore":6.4,"rationale":""},{"segmentIndex":82,"text":"without understanding. An LLM is a collaborator you can only trust by understanding what it did.","startTime":531.5799999999999,"endTime":538.645,"durationSeconds":7,"level":"C1","overallScore":6.6,"rationale":""},{"segmentIndex":86,"text":"AI generates code you didn't write, you kind of need it on day one. That's the actual shift. One","startTime":556.38,"endTime":564.2700000000001,"durationSeconds":8,"level":"B2","overallScore":6.4,"rationale":""},{"segmentIndex":89,"text":"it. We had to. Every senior I know, me included, built this systems thinking skill by failing","startTime":576.4599999999999,"endTime":583.5902941176471,"durationSeconds":7,"level":"B2","overallScore":6.4,"rationale":""}],"generatedAt":"2026-06-25T00:38:21.009Z","keyClipsTotalSec":574},{"videoId":"7zCsfe57tpU","chunkIndex":1,"totalChunks":3,"title":"Is this the only skill left? — Part 2 of 3","thumbnail":"https://i.ytimg.com/vi/7zCsfe57tpU/maxresdefault.jpg","duration":1330,"uploader":"Hak","youtubeUrl":"https://www.youtube.com/watch?v=7zCsfe57tpU","keywords":["ai-coding","software-engineering","systems-thinking","developer-productivity","career-growth","junior-developers","generalist","tech-trends"],"normalizedKeywords":["엔지니어링","커리어·성장","기술 트렌드"],"targetAudience":[{"who":"주니어 개발자","why":"AI 시대에 초반 학습과 실력 축적을 어떻게 해야 하는지 구체적 조언을 얻을 수 있음"},{"who":"시니어 개발자","why":"AI를 대체재가 아니라 판단력 증폭 도구로 쓰는 관점을 정리할 수 있음"},{"who":"비기술 창업자","why":"코드를 몰라도 시스템 사고로 AI 도구를 안전하게 활용하는 기준을 배울 수 있음"}],"normalizedAudience":["엔지니어·개발자","창업자·스타트업","학생·주니어"],"summary":"이 영상은 AI 코딩이 개발자의 일을 단순히 빨라지게만 하는 것이 아니라, 주니어가 실력을 쌓는 방식과 시니어가 가치를 내는 방식을 동시에 바꾸고 있다고 주장한다. 예전에는 막히면 코드와 씨름하며 배우는 과정 자체가 훈련이었지만, 이제는 AI가 그 'wrestle'을 없애면서 의도적으로 훈련하지 않으면 시스템 감각이 약해질 수 있다고 본다. 동시에 산업은 한때 줄였던 주니어 채용을 다시 늘리고 있으며, 결국 AI 시대에도 사람에게 필요한 것은 전체를 보고 판단하는 시스템 사고라는 점을 강조한다.\n\n후반부는 실천법에 집중한다. 주니어는 AI를 답을 빨리 얻는 도구가 아니라, 왜 그런지 묻고 손으로 다시 써보며 배우는 학습 파트너로 써야 하고, 시니어는 반복 작업은 AI에 맡기되 아키텍처와 큰 판단에 집중해야 한다. 비기술 창업자나 PM에게도 코드를 몰라도 시스템적으로 생각하고, 언제 전문가를 불러야 하는지 아는 능력이 중요하다고 말한다.","insights":["AI는 압박을 없앤 게 아니라 배우는 마찰을 없앴다.","주니어의 핵심 과제는 속도가 아니라 강제 훈련을 만드는 일이다.","AI 시대의 희소한 가치는 코드량이 아니라 시스템 판단력이다.","주니어는 AI로 이해를 깊게 하고, 시니어는 경험으로 검증해야 한다.","비기술자도 코딩보다 먼저 시스템 사고를 익혀야 한다."],"keyClips":[{"clipId":"7zCsfe57tpU:c1:1-18","startSegmentIndex":1,"endSegmentIndex":18,"startTime":603.5799999999999,"endTime":726.7277777777779,"durationSeconds":123.1,"preview":"주니어 채용의 역전","mustSee":false},{"clipId":"7zCsfe57tpU:c1:19-34","startSegmentIndex":19,"endSegmentIndex":34,"startTime":726.54,"endTime":836.4088888888889,"durationSeconds":109.9,"preview":"빠름보다 기준","mustSee":false},{"clipId":"7zCsfe57tpU:c1:35-53","startSegmentIndex":35,"endSegmentIndex":53,"startTime":836.3,"endTime":962.2878571428571,"durationSeconds":126,"preview":"시스템 사고의 본질","mustSee":true},{"clipId":"7zCsfe57tpU:c1:54-77","startSegmentIndex":54,"endSegmentIndex":77,"startTime":962.06,"endTime":1115.9360526315788,"durationSeconds":153.9,"preview":"직무별 AI 활용법","mustSee":false},{"clipId":"7zCsfe57tpU:c1:83-89","startSegmentIndex":83,"endSegmentIndex":89,"startTime":1149.9,"endTime":1201.78,"durationSeconds":51.9,"preview":"훈련하는 법","mustSee":false}],"curatedSegments":[{"segmentIndex":13,"text":"program. The code is just the shadow of it, as Peter Naur puts it. Fourty years later, today we","startTime":73.26,"endTime":80.27,"durationSeconds":7,"level":"B2","overallScore":7.6,"rationale":"비유가 강하고 핵심 메시지도 뚜렷."},{"segmentIndex":20,"text":"career asking \"if AI writes the code, how do we build the judgment to evaluate it? Should I","startTime":119.18,"endTime":125.79,"durationSeconds":7,"level":"B2","overallScore":7.8,"rationale":"학습자 고민을 직접 던지는 핵심 질문."},{"segmentIndex":38,"text":"how the parts fit together, when the strings should hold back, when the brass should come in.","startTime":241.26000000000002,"endTime":247.14794117647057,"durationSeconds":6,"level":"B2","overallScore":7.8,"rationale":"조율의 본질을 구체적으로 보여줌."},{"segmentIndex":55,"text":"If nothing tells you that, the system isn't working. Probably is pretending to work. Three,","startTime":356.38,"endTime":362.58,"durationSeconds":6,"level":"B2","overallScore":7.6,"rationale":"관측 없는 시스템의 위험을 날카롭게 짚음."},{"segmentIndex":57,"text":"the system in your head before you touch it? That's the theory Naur was talking about. Now,","startTime":369.26,"endTime":376.7535294117647,"durationSeconds":7,"level":"B2","overallScore":7.6,"rationale":"이론의 실천적 의미를 연결해 줌."},{"segmentIndex":65,"text":"there was a systems thinking failure, not a coding one. Where does state live? Well, everywhere all","startTime":423.34,"endTime":429.5938235294118,"durationSeconds":6,"level":"C1","overallScore":7.8,"rationale":""},{"segmentIndex":79,"text":"just a probabilistic translator. It's stochastic by nature. For the same input, you get different","startTime":513.18,"endTime":520.0236666666667,"durationSeconds":7,"level":"C1","overallScore":7.8,"rationale":"LLM의 본질을 대비적으로 설명한다."},{"segmentIndex":80,"text":"outputs almost each time. No guarantees. This compiler could introduce a security vulnerability,","startTime":519.8199999999999,"endTime":526.6507692307692,"durationSeconds":7,"level":"C1","overallScore":7.6,"rationale":"구체적 위험 예시로 설득력이 높다."},{"segmentIndex":17,"text":"now they need them back. So if you're a junior, you're not behind. You just got robbed of the","startTime":712.22,"endTime":717.2465789473686,"durationSeconds":5,"level":"B1","overallScore":6.8,"rationale":""},{"segmentIndex":26,"text":"one. And I think yes, there will be fewer, but the ones who will make it will be more valuable than","startTime":773.1,"endTime":782.3028571428572,"durationSeconds":9,"level":"B1","overallScore":6.8,"rationale":""},{"segmentIndex":59,"text":"great learning partner if you use it correctly. Yes, prompt, generate, but then study what comes","startTime":992.22,"endTime":998.7450000000001,"durationSeconds":7,"level":"B2","overallScore":7.8,"rationale":"AI 학습 활용법을 구체적으로 제안함."},{"segmentIndex":60,"text":"back. Treat the AI like an infinitely patient senior dev and ask it \"why this approach? What","startTime":998.62,"endTime":1005.5976470588237,"durationSeconds":7,"level":"C1","overallScore":8,"rationale":"질문 방식까지 제시한 고가치 조언."},{"segmentIndex":64,"text":"The next generation of seniors will be juniors who use AI to understand better and faster,","startTime":1022.9399999999999,"endTime":1028.325,"durationSeconds":5,"level":"B2","overallScore":7.4,"rationale":""},{"segmentIndex":11,"text":"the program. The program is what lives inside the programmer's head. How the pieces connect,","startTime":61.98,"endTime":67.68866666666666,"durationSeconds":6,"level":"B2","overallScore":6,"rationale":""},{"segmentIndex":64,"text":"error handling, but the app was apparently working until it didn't. Every single failure mode in","startTime":416.14,"endTime":423.39,"durationSeconds":7,"level":"B2","overallScore":6.6,"rationale":"시스템 취약성을 인상적으로 드러낸다."},{"segmentIndex":66,"text":"at once. Where does feedback live? Nowhere. We're blind. What breaks if I delete this? Never asked,","startTime":429.5,"endTime":436.73294117647055,"durationSeconds":7,"level":"B2","overallScore":7.2,"rationale":""},{"segmentIndex":78,"text":"it did the right thing. That's what makes it a trustworthy abstraction. An LLM is not that. It's","startTime":505.65999999999997,"endTime":513.2711111111112,"durationSeconds":8,"level":"C1","overallScore":6.4,"rationale":""},{"segmentIndex":82,"text":"without understanding. An LLM is a collaborator you can only trust by understanding what it did.","startTime":531.5799999999999,"endTime":538.645,"durationSeconds":7,"level":"C1","overallScore":6.6,"rationale":""},{"segmentIndex":86,"text":"AI generates code you didn't write, you kind of need it on day one. That's the actual shift. One","startTime":556.38,"endTime":564.2700000000001,"durationSeconds":8,"level":"B2","overallScore":6.4,"rationale":""},{"segmentIndex":89,"text":"it. We had to. Every senior I know, me included, built this systems thinking skill by failing","startTime":576.4599999999999,"endTime":583.5902941176471,"durationSeconds":7,"level":"B2","overallScore":6.4,"rationale":""}],"generatedAt":"2026-06-25T00:38:39.471Z","keyClipsTotalSec":574},{"videoId":"7zCsfe57tpU","chunkIndex":2,"totalChunks":3,"title":"Is this the only skill left? — Part 3 of 3","thumbnail":"https://i.ytimg.com/vi/7zCsfe57tpU/maxresdefault.jpg","duration":1330,"uploader":"Hak","youtubeUrl":"https://www.youtube.com/watch?v=7zCsfe57tpU","keywords":["ai-coding","software-engineering","spec-driven-development","system-design","developer-skills","code-review","ai-agents","technical-literacy"],"normalizedKeywords":["엔지니어링","기술 트렌드","커리어·성장"],"targetAudience":[{"who":"엔지니어","why":"AI 코딩 도구를 쓰면서도 시스템 사고와 코드 이해력을 유지하는 법을 배울 수 있음"},{"who":"주니어 개발자","why":"AI에 의존할수록 왜 기초 이론과 코드 읽기 훈련이 필요한지 알 수 있음"},{"who":"기술 리더","why":"팀이 AI를 쓸 때 무엇을 지켜야 장기적으로 실력이 무너지지 않는지 참고할 수 있음"}],"normalizedAudience":["엔지니어·개발자","학생·주니어"],"summary":"이 영상은 AI 코딩 에이전트를 효과적으로 쓰기 위해 필요한 핵심 역량이 무엇인지 정리한다. 화자는 AI가 코드를 대신 써주는 시대일수록 '무엇을 왜 만들고, 어떤 제약과 실패 모드가 있는지'를 먼저 정의하는 짧은 스펙이 중요하다고 말한다. 또한 생성된 코드를 그대로 받아들이지 말고, 삭제 테스트·대안 질문·수동 재작성 같은 습관으로 시스템 이해와 코드 읽기 근육을 유지해야 한다고 강조한다.\n\n결론적으로 AI는 개발자를 대체하기보다 타이핑을 대체할 뿐이며, 시스템으로 생각하는 능력은 여전히 사람의 영역이라고 주장한다. 다만 이 능력이 있으면 AI가 실력을 증폭시키고, 없으면 약점을 더 빨리 드러내기 때문에 주니어일수록 더 단단한 이론과 사고 훈련이 필요하다는 경고를 남긴다.","insights":["AI 시대의 핵심은 타이핑이 아니라 시스템 사고다.","짧은 스펙이 AI 협업의 가장 안전한 시작점이다.","삭제 테스트는 시스템 이해의 빈칸을 가장 빨리 드러낸다.","생성 코드를 검토·재작성해야 실력이 유지된다.","AI는 실력을 대체하지 않고 증폭하거나 노출한다."],"keyClips":[{"clipId":"7zCsfe57tpU:c2:1-4","startSegmentIndex":1,"endSegmentIndex":4,"startTime":1201.5800000000002,"endTime":1226.7933333333333,"durationSeconds":25.2,"preview":"스펙이 먼저다","mustSee":true},{"clipId":"7zCsfe57tpU:c2:5-7","startSegmentIndex":5,"endSegmentIndex":7,"startTime":1226.5400000000002,"endTime":1247.1208823529412,"durationSeconds":20.6,"preview":"삭제 테스트","mustSee":false},{"clipId":"7zCsfe57tpU:c2:8-10","startSegmentIndex":8,"endSegmentIndex":10,"startTime":1247.02,"endTime":1264.3283333333331,"durationSeconds":17.3,"preview":"코드 읽기 훈련","mustSee":false},{"clipId":"7zCsfe57tpU:c2:11-13","startSegmentIndex":11,"endSegmentIndex":13,"startTime":1264.22,"endTime":1285.1323333333335,"durationSeconds":20.9,"preview":"대체되는 건 타이핑","mustSee":true},{"clipId":"7zCsfe57tpU:c2:14-18","startSegmentIndex":14,"endSegmentIndex":18,"startTime":1284.8600000000001,"endTime":1315.3925000000002,"durationSeconds":30.5,"preview":"AI가 드러내는 실력","mustSee":true},{"clipId":"7zCsfe57tpU:c2:19-19","startSegmentIndex":19,"endSegmentIndex":19,"startTime":1315.1000000000001,"endTime":1321.11,"durationSeconds":6,"preview":"새 커리큘럼","mustSee":false}],"curatedSegments":[{"segmentIndex":13,"text":"program. The code is just the shadow of it, as Peter Naur puts it. Fourty years later, today we","startTime":73.26,"endTime":80.27,"durationSeconds":7,"level":"B2","overallScore":7.6,"rationale":"비유가 강하고 핵심 메시지도 뚜렷."},{"segmentIndex":20,"text":"career asking \"if AI writes the code, how do we build the judgment to evaluate it? Should I","startTime":119.18,"endTime":125.79,"durationSeconds":7,"level":"B2","overallScore":7.8,"rationale":"학습자 고민을 직접 던지는 핵심 질문."},{"segmentIndex":38,"text":"how the parts fit together, when the strings should hold back, when the brass should come in.","startTime":241.26000000000002,"endTime":247.14794117647057,"durationSeconds":6,"level":"B2","overallScore":7.8,"rationale":"조율의 본질을 구체적으로 보여줌."},{"segmentIndex":55,"text":"If nothing tells you that, the system isn't working. Probably is pretending to work. Three,","startTime":356.38,"endTime":362.58,"durationSeconds":6,"level":"B2","overallScore":7.6,"rationale":"관측 없는 시스템의 위험을 날카롭게 짚음."},{"segmentIndex":57,"text":"the system in your head before you touch it? That's the theory Naur was talking about. Now,","startTime":369.26,"endTime":376.7535294117647,"durationSeconds":7,"level":"B2","overallScore":7.6,"rationale":"이론의 실천적 의미를 연결해 줌."},{"segmentIndex":65,"text":"there was a systems thinking failure, not a coding one. Where does state live? Well, everywhere all","startTime":423.34,"endTime":429.5938235294118,"durationSeconds":6,"level":"C1","overallScore":7.8,"rationale":""},{"segmentIndex":79,"text":"just a probabilistic translator. It's stochastic by nature. For the same input, you get different","startTime":513.18,"endTime":520.0236666666667,"durationSeconds":7,"level":"C1","overallScore":7.8,"rationale":"LLM의 본질을 대비적으로 설명한다."},{"segmentIndex":80,"text":"outputs almost each time. No guarantees. This compiler could introduce a security vulnerability,","startTime":519.8199999999999,"endTime":526.6507692307692,"durationSeconds":7,"level":"C1","overallScore":7.6,"rationale":"구체적 위험 예시로 설득력이 높다."},{"segmentIndex":17,"text":"now they need them back. So if you're a junior, you're not behind. You just got robbed of the","startTime":712.22,"endTime":717.2465789473686,"durationSeconds":5,"level":"B1","overallScore":6.8,"rationale":""},{"segmentIndex":26,"text":"one. And I think yes, there will be fewer, but the ones who will make it will be more valuable than","startTime":773.1,"endTime":782.3028571428572,"durationSeconds":9,"level":"B1","overallScore":6.8,"rationale":""},{"segmentIndex":59,"text":"great learning partner if you use it correctly. Yes, prompt, generate, but then study what comes","startTime":992.22,"endTime":998.7450000000001,"durationSeconds":7,"level":"B2","overallScore":7.8,"rationale":"AI 학습 활용법을 구체적으로 제안함."},{"segmentIndex":60,"text":"back. Treat the AI like an infinitely patient senior dev and ask it \"why this approach? What","startTime":998.62,"endTime":1005.5976470588237,"durationSeconds":7,"level":"C1","overallScore":8,"rationale":"질문 방식까지 제시한 고가치 조언."},{"segmentIndex":64,"text":"The next generation of seniors will be juniors who use AI to understand better and faster,","startTime":1022.9399999999999,"endTime":1028.325,"durationSeconds":5,"level":"B2","overallScore":7.4,"rationale":""},{"segmentIndex":11,"text":"the program. The program is what lives inside the programmer's head. How the pieces connect,","startTime":61.98,"endTime":67.68866666666666,"durationSeconds":6,"level":"B2","overallScore":6,"rationale":""},{"segmentIndex":64,"text":"error handling, but the app was apparently working until it didn't. Every single failure mode in","startTime":416.14,"endTime":423.39,"durationSeconds":7,"level":"B2","overallScore":6.6,"rationale":"시스템 취약성을 인상적으로 드러낸다."},{"segmentIndex":66,"text":"at once. Where does feedback live? Nowhere. We're blind. What breaks if I delete this? Never asked,","startTime":429.5,"endTime":436.73294117647055,"durationSeconds":7,"level":"B2","overallScore":7.2,"rationale":""},{"segmentIndex":78,"text":"it did the right thing. That's what makes it a trustworthy abstraction. An LLM is not that. It's","startTime":505.65999999999997,"endTime":513.2711111111112,"durationSeconds":8,"level":"C1","overallScore":6.4,"rationale":""},{"segmentIndex":82,"text":"without understanding. An LLM is a collaborator you can only trust by understanding what it did.","startTime":531.5799999999999,"endTime":538.645,"durationSeconds":7,"level":"C1","overallScore":6.6,"rationale":""},{"segmentIndex":86,"text":"AI generates code you didn't write, you kind of need it on day one. That's the actual shift. One","startTime":556.38,"endTime":564.2700000000001,"durationSeconds":8,"level":"B2","overallScore":6.4,"rationale":""},{"segmentIndex":89,"text":"it. We had to. Every senior I know, me included, built this systems thinking skill by failing","startTime":576.4599999999999,"endTime":583.5902941176471,"durationSeconds":7,"level":"B2","overallScore":6.4,"rationale":""}],"generatedAt":"2026-06-25T00:39:04.248Z","keyClipsTotalSec":574}]}