Exactly. I've been disseminating a similar message for years.
The concentration of power in AI and the desire for control is by far the biggest danger of AI. It could lead to a few private companies and/or countries being in control of access to information, access to knowledge, and access to the tools of economic expansion.
It's a kind of medieval obscurantism akin to the Ottoman empire banning the use of the printing press for 200 years, in part to keep control of the dogma, but also to protect the corporation of the calligraphers and scribes.
Relevant historical bits about the Internet:
1. It took a deliberate decision by Al Gore and Bill Clinton to open up access of what was then ARPAnet to commercial entities and to the public, against the desires of the entrenched telecom industry. During a public roundtable about the "information superhighway" in 1993, the CEO of AT&T told Gore and Clinton "leave it to us". Gore said no.
2. In the late 1980s, setting up an Internet presence required buying proprietary hardware with proprietary OS and software stack from Sun Microsystems, HP, IBM, or Dell. By the 2000s, all of this was wiped out by commodity hardware, Linux, Apache, and an entirely free/open software stack. This migration to open platforms was the result of market forces.
Infrastructure wants to be open.
Foundation models are becoming an infrastructure and will inevitably become commoditized.
Long term, the money is in the application layer, which is what I, Arthur Mensch, Alex Karp, and others have been saying.
Free Book, Algorithms for Decision Making, 700 pages, CC license
https://t.co/Y1gFRYNfEb
本书全面介绍了不确定性决策算法
涵盖了与决策相关的各种主题
介绍了相关数学问题模型以及求解算法
本书算法采用 Julia 编写,也有项目将算法��译成 Python 代码
https://t.co/4PzzVVYOlK
一些大学开设了基于本书的课程:
Stanford University, CS238: Decision Making under Uncertainty
https://t.co/aMWA8xx0yM
University of Colorado, ASEN 5519-003: Decision Making under Uncertainty
https://t.co/yKdjiIlDuM
Brown University, CSCI 2951-F: Learning and Sequential Decision Making
https://t.co/yGBN1E1CP2
🚨New paper alert!
Context: KV cache has a boom in research -- 40x more papers on it in 2026 than 2024. But our experience shows that the industry can be reluctant because they are all *LOSSY* --- meaning the results with KV cache compression can be different from without it.
VeriCache solves exactly that problem without sacrificing the performance gain from the KV cache sparsification techniques.
Great work by the team led by Jiayi Yao (@EricYao43125144) !
🔥 Coding agents have become one of the hottest LLM workloads. But serving them looks nothing like serving a chatbot: 294× more input than output, hundreds of thousands of tool calls, and extremely long-tailed latency.
🚀 We are releasing the SyFI Coding Trace: ~4,300 real-world coding-agent sessions from our daily use, plus TraceLab, an open-source pipeline to collect, sanitize, analyze, and replay your own traces.
More in the thread below 🧵👇 (1/n)
《图解分布式系统原理》 (https://t.co/n3Bs9FBz5L)的英文版《Distributed System Illustrated》正式在海外的电子书发布平台leanpub上线 (https://t.co/CKxHlT9TMn)。在leanpub的书籍页面,有sample章节可以免费下载试读。
这意味着,从2024年底开始构思写作,历经一年多的打磨,这本教程终于做为一个完整的商品开始上架售卖。我知道现在这类技术书籍��定不好卖,但是我想看看,在我的努力下,最后能做到什么程度。小说《长安的荔枝》里,男主在面对困难时说的一句话:“既是退无可退,何不向前拼死一搏?”给了我很大的鼓励。这算是我个人独立从零开始完成的第一个作品。
纳瓦尔说:Learn to sell. Learn to build. If you can do both, you will be unstoppable.就这部作品而言,目前已经完成了build阶段,下面就是如何sell了。暂时的想法是在多写英文博客,以及这份教程会完全公开网页版,给leanpub购买链接引流。
另:中文电子版其实也已经完成,但是我暂时还没想好在哪个平台销售。
最近读到一篇很喜欢的文章:Zen and the Art of AI Research。
它最打动我的观点是:
做 AI research,很多时候拼的不是天赋,而是一种研究气质。
能不能长期泡在一个问题里,能不能接受实验失败和对好的结果保持怀疑,能不能在别人发了好 paper 之后,不只是焦虑,而是反过来问自己:我有没有在同样的深度上思考?
文章里有个很好的类比:做研究有点像打坐。
有灵感的时候,继续坐。
没有灵感的时候,也继续坐。
很多真正重要的想法,不是刷几篇热门论文就突然冒出来的,而是在反复阅读、动手构建、失败、debug、重新理解里慢慢长出来的。
这篇文章还提醒了一个很现实的问题:coding agents 会让研究跑得更快,但也会让人更容易失去对系统细节的掌控。
它可能帮你改了 prompt,缩短了 sequence length,换了 config,跑了一个看起来差不多、其实已经不一样的实验。
工程上这可能只是小问题,科研上就是大问题。
因为一个很小的改动,就可能改变整篇论文的结论。
所以 AI 时代的研究能力会���得更矛盾:
你要会用 AI 加速,但不能把理解外包给 AI。
你要跑得更快,但不能失去慢下来检查的能力。
你要拥抱 agents,但仍然要知道每个结果是怎么来的。
优秀的研究者需要在工具越来越强之后,依然能保持耐心、怀疑和清醒的人。
https://t.co/q5GLIVBUaZ
Final version of my book (with a new title)
Online Learning: A Modern Introduction Using
Convex Optimization
Especially proud of the Foreword by @NicoloCB!
It'll be printed by Cambridge University Press.
The end of 7 years of updates :)
https://t.co/NeqTSih2ra