I think that the superstar effect will only become more important moving forward. That's because lots more people will use AI, and each person will use AI systems much more heavily.
大多数人认为随着AI普及，薪酬差距可能会缩小或趋于稳定。但作者认为，随着AI用户数量和使用频率的增加，'超级明星效应'只会变得更加重要，顶级AI研究者的薪酬差距可能会进一步扩大，甚至出现1亿美元的年薪也不够的情况。
That matters because AI hype is dying down, and companies are shifting focus from buzzy pilots to deployment and integration, where cheaper and more customizable tools tend to win.
大多数人关注AI模型的性能和能力竞赛，但作者认为行业正从炒作阶段转向实际部署和集成，此时更便宜、可定制化的工具将获胜。这挑战了人们对AI发展重点的传统认知，表明中国开源模型的优势将在AI实际应用阶段更加凸显。
Some problems are open loop today but will close over time.
这一前瞻性观点暗示AI应用的发展轨迹是从开放循环到封闭循环的转变过程，这意味着当前许多需要人类判断的领域未来可能被AI完全自动化，具有深刻的战略意义。
While some experts have speculated that general models will win out in performance over specialized models—that scale and compute will beat curation—the success of these companies shows that the market is making a more nuanced bet.
市场正在形成一种更微妙的AI发展路径认知，表明通用模型与专业化模型可能在不同场景下各有优势。这种市场分歧暗示AI领域可能不会出现单一赢家，而是形成多元化发展格局。
70% of @Vercel's traffic is now coming from agents, up from 10% a year ago and on track to be 90% by end of year.
令人惊讶的是：AI代理在短短一年内从Vercel流量的10%激增到70%，预计年底将达到90%。这表明AI代理正在以前所未有的速度接管互联网流量，可能重塑我们使用网络的方式。
AI is moving faster than anyone predicted. When models change every 42 days, buyers can't assemble a best-of-breed stack.
令人惊讶的是：AI模型的更新速度如此之快，平均每42天就发生一次变化，这使得企业难以构建最佳组合的软件栈。这种极快的迭代速度彻底改变了传统的软件采购策略，迫使企业转向更全面的平台解决方案。
It is the only AI service to show a clear upward trend over this short time period
令人惊讶的是：在所有被调查的AI服务中，Claude是唯一显示出明确上升趋势的工具。这表明尽管市场竞争激烈，Claude仍然成功地在众多AI服务中脱颖而出，实现了持续增长。
Chinese models overtook their counterparts built in the U.S. in the summer of 2025 and subsequently widened the gap over their western counterparts.
令人惊讶的是：在短短几年内，中国开源语言模型生态系统已经全面超越美国，这标志着全球AI研发格局发生了重大转变。这一趋势不仅反映了中国在AI领域的快速进步，也暗示了未来技术领导力的可能转移。
Gemma points in the opposite direction: smaller models, local compute, more ownership.
大多数人认为AI发展必然走向更大、更集中的模型，但作者认为Google的Gemma 4代表了相反趋势。这挑战了AI发展的主流叙事，暗示未来AI可能分散到个人设备上，减少对大型基础设施的依赖，这与行业共识形成鲜明对比。
Mills, Anna, Maha Bali, and Lance Eaton. “How Do We Respond to Generative AI in Education? Open Educational Practices Give Us a Framework for an Ongoing Process.” Journal of Applied Learning and Teaching 6, no. 1 (June 11, 2023): 16–30. https://doi.org/10.37074/jalt.2023.6.1.34.
Annotation url: urn:x-pdf:bb16e6f65a326e4089ed46b15987c1e7
However, there is skepticism about AI’s ability to replace human teaching in activities such as judging writing style, and some have expressed concern that policy makers could use AI to justify replacing (young) human labor.
Maha describes here the primary concern I have with the pursuit of both AI and adaptive technologies in education. Not that the designers of such tools are attempting to replace human interaction, but that the spread of "robotic" educational tools will accelerate the drive to further reduce human-powered teaching and learning, leading perhaps to class-based divisions in educational experiences like Maha imagines here.
AI and adaptive tool designers often say that they are hoping their technologies will free up time for human teachers to focus on more impactful educational practices. However, we already see how technologies that reduce human labor often lead to further reductions the use of human teachers — not their increase. As Maha points out, that's a social and economic issue, not a technology issue. If we focus on building tools rather than revalorizing human-powered education, I fear we are accelerating the devaluation of education already taking place.