A

[Mark Kretschmann @mark_k]

Prediction 🔮: There will soon be a major security incident involving an open source AI model. It will be used as a pretext to demand a ban on open source AI.

This will be a staged false flag operation orchestrated by AI doomers to manufacture support for the restrictions they've wanted all along.

Don't fall for it.

[Patrick C Toulme @PatrickToulme]

$META I predict will imminently shutoff Anthropic and OpenAI model access for employees. This may very well affect the Anthropic and OpenAI IPOs.

Watermelon will be Fable class. This is 100% proof that Meta's experiment of having employees generate tens of thousands of RL envs works. There is no shortcut to doing this grunt work. (personally I think making RL envs is kinda fun)

Compute + Data + a ton of RL envs + some good people == Fable. There is no secret to frontier LLMs.

$META is obviously trying to drive down the price per token across the industry. I predict it will work.

[antirez @antirez]

Btw the Redis story repeats itself: I'm working at DwarfStar for free for the community and because I enjoy it. But I'm receiving criticisms, since people are worried that this will break their AI-richness plans. I want to say to everybody thinking that I should stop that each time you tell me this, I'll double down my efforts towards a completely no profit engine for local inference. Better to shut up basically.

[antirez @antirez]

Worth stressing.

[antirez @antirez]

If that happens Europe is officially fucked.

[@jason @Jason]

As I’ve warned, ChatGPT is coming for every single app

[Olivier Blanchard @ojblanchard1]

A must read (or must listen): https://t.co/2uGt1XhA8V . A dystopian but unfortunately surprising convincing of what could happen if Europe does not catch up on AI. It moved my prior. (and it is, despite the gloomy conclusions) a fun read.

[A @_ale_neri_]

@dom_gag_96 🍕

[Matthew Prince 🌥 @eastdakota]

Fascinating and clever.

[A @_ale_neri_]

@antirez Have you thought of supporting the new open source model from nvidia Nemotron 3 Ultra?

[A @_ale_neri_]

@lucasmeijer @badlogicgames agent-browser cli from vercel is nice

[antirez @antirez]

Welcome to DS4, a specialized inference engine for DeepSeek v4 Flash. https://t.co/UrUJz5I2R1

This project would have been impossible without the existence of llama.cpp and GGML and the work of @ggerganov and all the other contributors. Thanks!

[Mario Zechner @badlogicgames]

extremely happy that we are in q2 2026, and engineers i look up to are plowing a path for the local model future.

yes, we are still beholden to some lab publishing weights. but i take that over companies with questionable motives and business practices.

[Steren @steren]

Announcing Cloud Run sandboxes:

Secure on-the-fly code execution: Spin up ephemeral, isolated sandboxes from within Cloud Run resources.

Safely execute agent-generated code, scripts, or Chromium.

[Andrej Karpathy @karpathy]

Something I've been thinking about - I am bullish on people (empowered by AI) increasing the visibility, legibility and accountability of their governments.

Historically, it is the governments that act to make society legible (e.g. "Seeing like a state" is the common reference), but with AI, society can dramatically improve its ability to do this in reverse. Government accountability has not been constrained by access (the various branches of government publish an enormous amount of data), it has been constrained by intelligence - the ability to process a lot of raw data, combine it with domain expertise and derive insights. As an example, the 4000-page omnibus bill is "transparent" in principle and in a legal sense, but certainly not in a practical sense for most people. There's a lot more like it: laws, spending bills, federal budgets, freedom of information act responses, lobbying disclosures... Only a few highly trained professionals (investigative journalists) could historically process this information. This bottleneck might dissolve - not only are the professionals further empowered, but a lot more people can participate.

Some examples to be precise: Detailed accounting of spending and budgets, diff tracking of legislation, individual voting trends w.r.t. stated positions or speeches, lobbying and influence (e.g. graph of lobbyist -> firm -> client -> legislator -> committee -> vote -> regulation), procurement and contracting, regulatory capture warning lights, judicial and legal patterns, campaign finance... Local governments might be even more interesting because the governed population is smaller so there is less national coverage: city council meetings, decisions around zoning, policing, schools, utilities...

Certainly, the same tools can easily cut the other way and it's worth being very mindful of that, but I lean optimistic overall that added participation, transparency and accountability will improve democratic, free societies.

(the quoted tweet is half-ish related, but inspired me to post some recent thoughts)

[flavio @flaviocopes]

How Axios was compromised 🤯

[Marktechpost AI @Marktechpost]

Yann LeCun’s New LeWorldModel (LeWM) Research Targets JEPA Collapse in Pixel-Based Predictive World Modeling

Predictive world models often 'cheat' via representation collapse. Yann LeCun’s team introduced LeWorldModel (LeWM), the first JEPA to train stably end-to-end from pixels without heuristics like stop-gradients or EMA.

LeWM utilizes a streamlined two-term objective featuring SIGReg. By enforcing Gaussian-distributed latents via the Cramér-Wold theorem, it prevents collapse while capturing meaningful physical structure.

Efficiency: Uses ~200× fewer tokens than DINO-WM, enabling 48× faster planning (0.98s vs 47s).....

Full analysis: https://t.co/4aHug8Jl29

Paper: https://t.co/zqMMrMhORl

Repo: https://t.co/hmGg1e6mYF

Website: https://t.co/THtM1fhrEX

@ylecun

[Guillermo Rauch @rauchg]

https://t.co/DduIC3wkWT

[templar @tplr_ai]

We just completed the largest decentralised LLM pre-training run in history: Covenant-72B. Permissionless, on Bittensor subnet 3.

72B parameters. ~1.1T tokens. Commodity internet. No centralized cluster. No whitelist. Anyone with GPUs could join or leave freely.

1/n

[sui @birdabo]

Gemini 3.1 + Claude 4.6 seems like a dangerous combo.

a dev built a geospatial tracking system with live planes, satellites traffic cams, and panoptic detection.

this is literally a CIA tool lmao.

i guess this is what happens when AI gets too good