In retrospect, it will have been obvious what's about to happen. Take your #Bitcoin    and other crypto off exchanges soon.

User funds are safe! It does look like a hack on Binance. Appreciate the quick freeze from Tether (~$971m) and Circle (~$823m). I don't manage Binance (.com) anymore, but as a shareholder, I expect all user funds will be fully covered.

POLYMARKET USERS REPORT THAT FRONTEND MAY BE COMPROMISED, AND ASSETS ARE BEING WITHDRAWN WITHOUT PERMISSION: DISCORD

Download these before the internet gets restricted during the cyber pandemic github.com/BaaRai/ARTEX github.com/deepseek-ai/DeepI… github.com/nationalsecuritya… huggingface.co/dealignai/GLM… github.com/yifanzhang-pro/re… github.com/wheresryan22/anat… github.com/Gen-Verse/RSI-Any… github.com/solidsf-inc/solid… github.com/google-research/h… huggingface.co/GroundingPI/G… huggingface.co/datasets/sece… (figure out if u need raw, papers, pdf version etc based on storage) Genome models • Evo 2 (40B): huggingface.co/arcinstitute/… • Evo 2 (7B): huggingface.co/arcinstitute/… • Genos-1.2B: huggingface.co/BGI-HangzhouA… • Genos-10B: huggingface.co/BGI-HangzhouA… Pathology models • OpenMidnight (Sophont): huggingface.co/SophontAI/Ope… • GenBio-PathFM: huggingface.co/genbio-ai/gen… Protein design • BoltzGen: huggingface.co/boltzgen/bolt… • Pinal (text-to-protein): huggingface.co/westlake-repl… huggingface.co/dealignai/Dee… orcarouter/Qwen3.8-27B-Uncensored-GGUF huggingface/carbon ESM-2 + ESMFold OpenFold PubMed Central Open Access github.com/cloudflare/securi… github.com/antoniaci/blackbi… github.com/soxoj/maigret github.com/smicallef/spiderf… github.com/laramies/theHarve… github.com/achillean/shodan-… Also buy a hack rf one

For the second week in a row, LINK has been the most traded asset on Coinbase for Agents. The top 3 assets by notional trading volume from 10/19 to 10/25: 🥇 $LINK 🥈 $BTC 🥉 $ETH Coinbase for Agents is now natively available on Grok, Grok Bot, Perplexity Computer, Composio, and via MCP on other AI platforms.

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…

The imbalance between what attackers can do with abliterated models (with no KYC btw) and what you are allowed to do within cyber constraints is insane. I have done a full-face scan and passport upload to Anthropic, and Claude won't even do basic defence. Feels very problematic