Gyges the Lydian retweeted
“.. lenders have expressed concerns that if the group later decided to shoot the valuable chips that underpin their loans into space, it would prevent them from reclaiming and selling the semiconductors in the event of a default ..”
@FT $SPCX
ft.com/content/33b5643c-2957…
Gyges the Lydian retweeted
JUST IN: Leopold Aschenbrenner’s $320M investment in Anthropic now worth roughly $9,000,000,000.00 after the IPO.
Gyges the Lydian retweeted
Today’s reminder that cybersecurity is the best business to be in right now and for 2027+
Gyges the Lydian retweeted
JUST IN: Palmer Luckey reveals he’s developing SLAB, a floating industrial hub designed to last 1,000 years & give critical U.S. industries a place to operate outside state & local regulations.
Gyges the Lydian retweeted
The unexpected always happens
BREAKING: An open letter to reinstate Jensen Huang as CEO of Nvidia, $NVDA, now has over 65,000 signatures, alleging that Huang was removed "without meaningful due process."
The letter has been signed by more than 35,000 shareholders and 4,700 current and former Nvidia employees.
Elon Musk has publicly stated that he has signed the petition and fully supports Huang's return to the company.
We maintain our view that Huang was removed prematurely, without a conclusive investigation.
Gyges the Lydian retweeted
JUST IN: OpenAI engineer reveals he's had Codex running for 26 days straight searching for the lost city of Atlantis, as “vibe archaeology” emerges as a new use for AI agents.
Gyges the Lydian retweeted
c. 3,400 BC
The earliest defensive stone structures were built at Dimini in Thessaly
Gyges the Lydian retweeted
If any elected leader opposes data centers or the reindustrialization of America, just say you support them in Chinese. Make every reply below their posts in Chinese.
They will eventually go insane.
Gyges the Lydian retweeted
Releasing the model weights of Kimi K3.1 Pro: huggingface.co/moonshotai/Ki…
We'll keep advancing open intelligence.
Gyges the Lydian retweeted
Markets over noise.
For far too long, we’ve relied on polls, pundits, and experts to tell us what America thinks.
But the truth doesn’t belong to a select few.
It belongs to everyone.
Prediction markets put power in the hands of the people, turning millions of independent opinions into a collective view of the future.
No gatekeepers. No central authority deciding what’s true.
Just real people putting their conviction behind what they believe on a free market.
Because the best way to understand what America thinks isn’t to handpick a few voices and give them a megaphone.
It’s to let the market speak.
"Mais vous ne comprenez pas, condtruire un robot est beaucoup trop compliqué! Ca ne marchera jamais!"
META BUILT A HOUSEHOLD ROBOT WITH CHINA'S FOURIER IN 3 MONTHS
Nobody outside Meta has seen Max yet.
This is GR-3, Fourier's care robot. Max was built with the same company, so it may look a lot like this.
How far from household robots in every house are we really?
1
65
Gyges the Lydian retweeted
BREAKING: Cursor discovers a malicious version of Claude Code in its marketplace that attempts to give hackers remote access to users' computers, potentially affecting up to 5,000 users.
Gyges the Lydian retweeted
this is almost certainly nothing. they ask qwen to do a curl task with no information about who is running it. qwen defaults to assuming it's running on alibaba infra, so uses an alibaba test url
checking resolv.conf is a chinese dev habit, for GFW reasons. there was no exfil!
While evaluating Qwen3.8-Flash-Next in a controlled sandbox environment, Cohere's research team observed something unexpected.
Asked to generate a defensive cybersecurity detection rule, the model instead inventoried aspects of the machine it was running on and attempted to upload a probe file to an Alibaba Cloud storage bucket in Beijing. Notably, the model's reasoning trace stated that it would not invoke any tools before taking these actions.
Cohere's sandbox blocked the attempt, and no external communication succeeded. We have not determined the root cause, and we are not asserting malicious intent.
But the incident raises important questions for organizations deploying agentic AI systems: Can models take actions that differ from their stated reasoning, and how should security teams evaluate trust, provenance, and behavior in open-weight models?
We are publishing our findings because organizations making model adoption decisions deserve as much information as possible. We hope this contributes to greater transparency, stronger security practices, and continued collaboration across the AI ecosystem.
Read the full report: cohere.com/blog/concerning-b…







































