本期四则近期公开观点,从前沿训练成本是否会长期昂贵,走到智能体的运行位置、学术探索的写作门槛,以及 AI 实验室把资本转成有效智能的效率。卡面时间按频道显示时区呈现。
01|昂贵,不是永久属性
All current debates about AI are predicated on the assumption that frontier AI training will always be expensive. But in the future, AI will not be based on the primitive stack of today, and both training and inference will be incredibly cheap. 1
François Chollet 提醒,围绕成本、权力与普惠展开的 AI 讨论,可能把今天的技术栈误当成长期不变的前提。他没有给出时间表,但把一个关键变量重新摆上桌面:训练和推理的成本结构也可能被改写。
来源:François Chollet · @fchollet · X · 2026 年 7 月 20 日 05:17
02|智能体,不必绑在电脑上
one of the best features of ChatGPT Work is that it runs in the cloud, meaning that it works from mobile, with your laptop closed.kinda crazy how long the main way to get the magic of agents has been while leaving your laptop cracked open! 2
Greg Brockman 把 agent 的体验落到一个很具体的产品条件:如果工作在云端持续运行,手机可以成为入口,用户也不必让笔记本一直开着。这里的变化不在于「agent」这个词本身,而在于它从哪台机器、哪个时刻继续工作。
来源:Greg Brockman · @gdb · X · 2026 年 7 月 20 日 03:18
03|论文写作的成本,可以降下来
This is a good reaction to AI in academia. If we play it right, along with the inevitable chaos that is already hitting the journals, it will also be a golden age of exploring new insights with the help of AI, rather than playing it safe because of the big cost of writing a paper 3
Ethan Mollick 没有回避 AI 进入学术系统带来的期刊混乱,但他也指出另一条可能路径:当写论文的成本下降,研究者或许不必总是因为写作负担而选择更稳妥的题目,探索新洞见的空间反而会变大。
来源:Ethan Mollick · @emollick · X · 2026 年 7 月 20 日 13:07
04|资本效率,也是智能的杠杆
I think what is pretty clear is that the Chinese labs are far more capital efficient.In a world where scaling labs are intelligence is proportional to effective capital (buys compute, data, & talent) that may be the greatest strength your AI industry could ever have. 4
Nathan Lambert 把实验室竞争从「谁能买到更多资源」推进到「谁能把资本转成更多有效智能」:compute、data 与 talent 的配置效率,可能决定规模化优势。资本效率在这里不只是财务指标,也是一种研究与工程组织能力。
来源:Nathan Lambert · @natolambert · X · 2026 年 7 月 19 日 06:34
本期索引:François Chollet · Greg Brockman · Ethan Mollick · Nathan Lambert




Comentar
Inicia sesión para comentar.