https://www.youngju.dev/blog 천천히 올바르게. AI Researcher & DevOps Engineer Youngju's blog. GPU/CUDA, LLM, MLOps, Kubernetes AI workloads, and data engineering — plus mindset essays on confidence, routines, health, and sport psychology. ko fjvbn2003@gmail.com (Youngju Kim) fjvbn2003@gmail.com (Youngju Kim) Sat, 15 Aug 2026 00:00:00 GMT https://www.youngju.dev/blog/career/2026-08-15-career-skills-working-with-ai-tools.en https://www.youngju.dev/blog/career/2026-08-15-career-skills-working-with-ai-tools.en Saying someone is good with AI tools carries no information until you know what they delegated and how they judged the result. This piece covers the criterion that draws the delegation line by verification cost rather than difficulty, the four-line delegation card to write before handing work over, the procedure that changes the order in which you read a result, and four patterns where productivity goes the wrong way. It cites a randomized controlled trial and a large developer survey, and it states what those numbers do not say. No product talk. Part 8 of the What Stays Expensive series. Sat, 15 Aug 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) careerskillsaiverificationengineering-practice https://www.youngju.dev/blog/career/2026-08-15-career-skills-working-with-ai-tools.ja https://www.youngju.dev/blog/career/2026-08-15-career-skills-working-with-ai-tools.ja AIツールをうまく使う、という言い方は、何を委ね、結果をどう判定したかを抜くと何も意味しません。この記事は、委譲の可否を難易度ではなく検証費用で分ける基準、委ねる前に四行で書く委譲カード、結果を受け取ったあと読む順序を変える検証手順、そして生産性がむしろ落ちる四つのパターンを扱います。無作為化対照試験と大規模な開発者調査の数字を引きつつ、その数字が何を語っていないかも書きました。特定の製品の話はしません。高いまま残る技術シリーズの第8回です。 Sat, 15 Aug 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) careerskillsaiverificationengineering-practice https://www.youngju.dev/blog/career/2026-08-15-career-skills-working-with-ai-tools https://www.youngju.dev/blog/career/2026-08-15-career-skills-working-with-ai-tools AI 도구를 잘 쓴다는 말은 무엇을 맡겼고 결과를 어떻게 판정했는지를 빼면 아무것도 뜻하지 않습니다. 이 글은 위임 가능성을 난이도가 아니라 검증 비용으로 가르는 기준, 맡기기 전에 네 줄로 적는 위임 카드, 결과를 받은 뒤 읽는 순서를 바꾸는 검증 절차, 그리고 생산성이 오히려 떨어지는 네 가지 패턴을 다룹니다. 무작위 대조 실험과 대규모 개발자 설문에서 나온 숫자를 인용하되 그 숫자가 무엇을 말하지 않는지도 함께 적었습니다. 특정 제품 이야기는 하지 않습니다. 비싸게 남는 기술 시리즈의 8편입니다. Sat, 15 Aug 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) careerskillsaiverificationengineering-practice