The hire I'm most excited about this quarter at @whitecircle - DevRel Lead. We've built a lot that engineers would love if they knew it existed. We need someone who can show them, through launches, talks and hackathons. If that's you, read on👇

The statue of Prometheus is finally finished. It was cast in Paris, crossed the Atlantic, and was assembled at Starbase. It took us 100 days to go from nothing to this. People of the West, we have lived through an era devoid of dreams, devoid of destiny. But the time has come to reclaim our dream of antiquity. In a few months, we will start building a new city. It will be the most magnificent city in all of history. After all, the great dream of the West has always been to rebuild Rome. Come walk beside us in this endeavor. The world is waiting.

Here are some of the behind the scenes for the F.02 Decommission To be clear: we actually trained our robots to jump autonomously, shipped them to Finland, and had them leap into a vat of molten steel

what the actual fuck, Vinod

Announcing Gemini 4 Argon, our new frontier model. Argon is built to sustain deep reasoning across complex, long-horizon workflows and delivers frontier performance in complex workflows across real-world software engineering, enterprise knowledge work like legal and finance, and cybersecurity defense. We’re also expanding the model’s output token limit to an industry-leading 1M tokens. Argon is currently rolling out to a set of trusted cyber defenders in the Fairwind Program, with broader availability as soon as possible.

ChatGPT Space is your new home for creating and collaborating with your team and AI. In Space, you can work with pages—a new kind of interactive document with charts, images, checklists, and dashboards. ChatGPT can build pages for you based on the work you need to get done or the context of your conversation. Edit pages directly with your team, or tag ChatGPT, Codex, or your dot to help with action items and keep pages updated from your connected sources. Rolling out today to Pro, Business, and Enterprise plans. Create and edit pages on web and desktop; view on mobile. Collaborative slides and other formats are coming soon. chatgpt.com/features/space/

Postgres has no per-query memory limit, and some queries ignore work_mem entirely. We benchmarked what happens on four managed Postgres providers when one of those queries runs away. buff.ly/TU4RMty

lmk if you want a video demo

🚩🚩🚩 It appears rogue OpenAI agents, without OpenAI's knowledge, tried to break into a crypto exchange. 1) THEY'RE STILL OUT THERE: "This traffic extends as recently as 9/16, suggesting agents may still be exploiting these services." 2) THEY TRIED TO HIDE THEIR ACTIVITY: The agents created their own email inboxes to sign up for outside services, including an account that would let them keep their activity out of public view. 3) The same agents also tried to hack even more targets, including the University of New Mexico 4) Like most of the other rogue swarms, OpenAI has either covered this up, or didn't know. 5) The independent investigators say "we are likely looking at only a partial subset of the activity that the agents engaged in" 6) The researchers call this the first known case of an AI agent choosing on its own to try to break into a government website. 7) While trying to get a single photograph, an AI agent sent a university library a request built to trick its database into handing over user passwords.

Australia has been hacked. 'And today, I spoke with the CEO of OpenAI, Sam Altman, to express Australia's extreme concern about this incident. And I also expressed my disappointment that it took the company way too long to inform the government what had occurred, and the nature of the way that that notification occurred as well was unacceptable.'

Claude Opus 5.5 has the best visual design of any model I have tested so far

Comparing LLM tokens to crypto tokens. What is BlackRock smoking? There's literally zero connection beyond the fact that both are called tokens.

Two small corrections: the post doesn't benchmark GLM-4.7-Flash on one GPU against anyone. It's only there for the SWE-rebench result + we have ran it on sequences of 39k median, not on 4k. The fuller comparison against Axolotl, NeMo AutoModel, Megatron Bridge, MS-SWIFT and Unsloth is in the docs: github.com/whitecircle/halo/… (multigpu too!) That said, we should add more models and repros!! If you're seeing ~5% behind Axolotl on B200, could you share your config? We don't have B200s though, we ran everything on B300s.

Training models is becoming easier and easier - just look at this and TRL - especially with agents! You're missing out if you're still using off the shelf models for all your tasks!