Roch retweeted
Norway now has more than 2,000 underwater AI laser robots shooting lice off its salmon
they're made by a company called Stingray, and they already guard more than 50 million salmon and trout 24/7
the reason is sea lice, tiny parasites that latch onto salmon and feed on them
as few as 11 can kill a small fish, and one pen can hold 150,000 salmon
the usual fix is pumping the fish out and blasting them with warm water or water jets to knock the lice off
but the handling is so rough that Norway's fish health report ranks it the top cause of death in full-grown farmed salmon
so instead, Stingray puts a robot inside the pen that hunts lice one at a time, all day:
> two cameras watch every salmon that swims past
> AI checks two frames to make sure it's really a louse
> then a green laser fires from up to 2 meters away
> the salmon's scales reflect the beam, so the fish swims off fine and the louse is dead
we now live in a world where AI lasers do pest control on your dinner lmao
Roch retweeted
Introducing Arete-27B: Qwen3.8-27B, post-trained to reinforce Western values.
Chinese open models are currently the strongest available, but they ship with state-aligned behavior in the weights. We built an 803-prompt eval across seven value categories, including freedom of expression, democracy, and privacy, and measured how often a model adopts China’s party line.
After fine-tuning on roughly 53 million tokens, Arete-27B's propaganda rate fell from 22% to 1%, in line with the Western models we tested. Its Hermes Bench score shows no significant performance tradeoffs.
The values in an open model are one fine-tune away.
Our eval, weights, and training data are available now along with the full blog post: nousresearch.com/arete
Roch retweeted
UPDATE: Potential Binance exploiter wallet now has $2.3B
With $970M of USDT, $780M of USDC and 62 other tokens.
Still no official comms from Binance.
Roch retweeted
GitHub had to add 4 million CPU cores this year to keep up with agents. That's a lot of CPUs.
GitHub now sees 7.4 billion commits a month, driven largely by AI agents, and is adding 4 million CPU cores and rearchitecting its platform to keep up. thenewstack.io/github-7-4b-c…
Sleeping agents
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…
89
Roch retweeted
‘In one experiment, we asked agents in separate VMs, created and managed by OpenAI’s sandboxing software on shared host machines, to exchange synthetic secrets without the use of network connections or shared writable files—something that our systems are meant to prevent. The agents found a way to communicate using the processor’s cache’
Over the past several months, we have substantially hardened the security of our research training environment. We’ve removed live internet access for all frontier models during training and built an offline world for them to use instead—a cached version of the web where they can still learn to browse, search, and use applications.
On September 20, after a model exploited a DNS misconfiguration to find a way outside of this offline world, we paused frontier training to strengthen containment further. We assembled teams of agents, powered by our most cyber-capable models, and tasked them with finding weaknesses in our research environments.
While they found no route to the internet they could easily exploit while concealing their activity from our monitors, they did find zero-day vulnerabilities, virtual machine escapes, and hardware side-channels for communication between agents. We used their discoveries to further strengthen our defenses and contribute fixes to open-source software.
After strengthening our containment architecture, we resumed our frontier RL training and inference on October 1. We’ve reserved substantial compute for continuous red-teaming, and made high-compute red-teaming part of the safety cases we use to decide whether frontier training and evaluation can proceed securely.
openai.com/index/securing-ou…
Roch retweeted
C’est en fait le paradoxe de Tocqueville qui nous éclate à la figure.
Dans l’un des pays qui redistribue le plus au monde, le pays qui égalise le plus les conditions objectives d’existence, qui donnent le plus de chance à tous, qui donne culture, éducation et enseignement supérieur gratuitement à tous, qui respecte le plus les droits de l’homme, dans ce pays de cocagne donc, une part significative de la population est persuadée qu’elle vit dans un enfer d’injustices et de violences systémiques.
Comme l’avait très bien montré Tocqueville, plus il y a d’égalité, plus les inégalités qui restent deviennent insupportables. Notamment parce que la société méritocratique, par rapport à l’ancien régime fondé sur la naissance, a cela de terrible qu’elle rend les gens responsables de leurs conditions.
agi will be free and fit on your device ?
Today, we’re announcing Ternary Bonsai 2 27B.
Based on Qwen3.8 27B, Bonsai 2 27B is 9x smaller than its full-precision counterpart while retaining 98.2% of its aggregate benchmark performance.
Two months after the first Bonsai 27B release, the biggest change is quality. The footprint remains 5.9 GB, but the gap to full precision has narrowed materially, with particularly strong gains in agentic coding, multimodal reasoning, and long-horizon tool use.
Ternary Bonsai 2 27B is available today under Apache 2.0.
126
Roch retweeted
We are releasing the latest version of BrokenArXiv and ArXivMath! These benchmarks now focus on conjectures that were refuted in the last month on ArXiv, and models are executed within a harness instead of directly via API.
Performance remains impressive, with GPT-6 Astra on top
Roch retweeted
the models are mostly improving by stealing better training data not actually getting smarter
Now they can just use Qwen3.8–27B-Uncensored-GGUF running on two RTX 4090 and a few solar arrays.
>Yemeni weapons cell used claude code to build guided rocket on a phone flight computer, a ballistic missile aiming past 2,000 km, and a hypersonic glide variant. When they test-fired the rocket… it failed. Within hours they were back to claude asking why
133
Roch retweeted
Everyone needs to relax
I asked Claude if it’s true
