別再讓 AI 代打了！練習進步 48%、考試反而輸 17%，新人要怎麼變高手？

Anthropic 一位工程主管這個月宣布：他的團隊不考程式了，因為那種小型程式題，Claude 已經寫得比大多數應徵者好。史丹佛的研究也發現，二十二到二十五歲、在最容易被 AI 接手的工作裡，年輕人的就業比該有的水準少了百分之十九。 那些難的東西，沒有人一開始就會。每個高手都是從最簡單的工作爬上去的，而 AI 正在把最下面那幾階拿走。 這集拆開「新手階梯」：一九七零年代裁縫學徒為什麼先學縫釦子；為什麼 AI 最幫新手，卻可能讓你以為自己學會了（「代打」）；一九八三年一篇論文怎麼預言了今天；AlphaGo 之後棋士為什麼反而變強；還有你可以自己把階梯蓋回來的四個方法。 你現在有沒有一件事，是 AI 幫你做得很好，但你其實自己不會的？留言告訴我。 主持：溪魚（千白頻道的科技／科普節目） 章節 00:00 AI 比新人還會寫程式，然後呢？ 01:57 一、新手階梯 04:36 二、AI 最幫的，是新手 05:59 三、代打 08:37 四、一九八三年的預言 10:53 五、反過來想：AlphaGo 之後 13:01 六、能力長在卡關的地方 14:18 七、自己蓋階梯的四個方法 16:32 結尾：學唱歌的小鳥 參考資料 Felix Rieseberg, "Hunger beats tenure: advice for junior engineers"（2026-10-27）：https://felixrieseberg.com/hunger-bea... Brynjolfsson, Chandar & Chen, "Canaries in the Coal Mine? Six Facts About the Recent Employment Effects of Artificial Intelligence," Stanford Digital Economy Lab（2026-08 修訂）：https://digitaleconomy.stanford.edu/n... Anthropic, "How employment is changing at Anthropic"（2026-10-16）：https://www.anthropic.com/research/em... Lave & Wenger,《Situated Learning: Legitimate Peripheral Participation》, Cambridge University Press, 1991 Brynjolfsson, Li & Raymond, "Generative AI at Work," Quarterly Journal of Economics 140(2), 2025：https://academic.oup.com/qje/article/... Bastani et al., "Generative AI without guardrails can harm learning: Evidence from high school mathematics," PNAS 122(26), 2025：https://doi.org/10.1073/pnas.2422633122 Shen & Tamkin（Anthropic）, "How AI Impacts Skill Formation"（2026）：https://arxiv.org/abs/2601.20245 ；https://www.anthropic.com/research/AI... Anthropic, "How AI is transforming work at Anthropic"（2025-12-02）：https://www.anthropic.com/research/ho... Bainbridge, "Ironies of Automation," Automatica 19(6), 1983 Shin, Kim, van Opheusden & Griffiths, "Superhuman artificial intelligence can improve human decision-making by increasing novelty," PNAS 120(12), 2023：https://arxiv.org/abs/2303.07462 Kornell, Hays & Bjork, "Unsuccessful retrieval attempts enhance subsequent learning," Journal of Experimental Psychology: Learning, Memory, and Cognition 35(4), 2009 Buçinca, Malaya & Gajos, "To Trust or to Think: Cognitive Forcing Functions Can Reduce Overreliance on AI in AI-assisted Decision-making," CSCW 2021：https://arxiv.org/abs/2102.09692 Konishi, "The role of auditory feedback in the control of vocalization in the white-crowned sparrow," Zeitschrift für Tierpsychologie 22, 1965 Hayase et al., "Vocal practice regulates singing activity–dependent genes underlying age-independent vocal learning in songbirds," PLoS Biology, 2018：https://pmc.ncbi.nlm.nih.gov/articles...