DeepMind has hired a philosopher to study AI consciousness. I'm really curious about what exactly they've observed internally.
SAFETY
-
Chinese AI Robot Transforms Hazardous Grain Warehouse Safety Standards
By
–
Chinese #AI-Powered #Robot Revolutionizing Safety in Hazardous Grain #Warehouses
— Ronald van Loon (@Ronald_vanLoon) 14 avril 2026
by @tweetciiiim
#Robotics #Engineering #ArtificialIntelligence #Innovation #Technology pic.twitter.com/trQkCBKkY1Chinese #AI-Powered #Robot Revolutionizing Safety in Hazardous Grain #Warehouses
by @tweetciiiim #Robotics #Engineering #ArtificialIntelligence #Innovation #Technology -
RIFT: Taxonomy of Rubric Failure Modes for AI Evaluation
By
–
Rubrics have become widely accepted for evaluating agents and models, but how are we evaluating the rubrics themselves? In a new paper we’ll be presenting at the Data-FM workshop at @iclr_conf
, we introduce RIFT: a taxonomy of 8 rubric failure modes across: ➜ reliability
➜ -
AI Index 2026: Breakthrough Capabilities and Critical Challenges Ahead
By
–
The #AIIndex2026 report reveals a field hitting breakthrough capabilities, while raising urgent questions about environmental costs, transparency, and who benefits from the technology. Read the main findings here https://
hai.stanford.edu/news/inside-th
e-ai-index-12-takeaways-from-the-2026-report
… -
Malicious Intermediary Attacks on LLM Agent Supply Chain Security
By
–
"Your Agent Is Mine: Measuring Malicious Intermediary Attacks on the LLM Supply Chain" The danger of agent security can also exist in the API router, that is between the agent and the provider. As these routers can read prompts, keys, and tool calls in plaintext, even rewrite
-
DeepAgents introduces filesystem permissions for secure resource management
By
–
🔒 new in deepagents: filesystem permissions shared resources and org-wide policies are exactly the kind of files you want your agent to read but never overwrite. filesystem permissions let you enforce that with simple declarative allow/deny rules!
→ View original post on X — @langchain, 2026-04-13 19:16 UTC
-
Exploit Costs More Than Human Security Researchers
By
–
-
Anthropic and OpenAI: Decoding Strategic AI Model Names
By
–
At face value, the claim sounds like a safety warning. It also seems to signal something unusually capable, consequential and maybe beyond what rivals have. It’s hard not to notice the names @AnthropicAI and @OpenAI are using:
— "Spud" (still just a codename) calls to mind -
Anthropic Mythos Sets New AI Competition Bar for OpenAI
By
–
Now that @AnthropicAI has introduced Mythos, the bar is set for @OpenAI and “Spud.” In today's @BigTechnology newsletter, @Kantrowitz look at what comes next. We also ask series of questions — like whether “too dangerous” is a new trend that's part warning & part marketing.
-
UK AI Security Institute Evaluates Claude Mythos Preview Safety
By
–
Very interesting evaluation from the UK’s AI Security Institute of the not yet publicly available Claude Mythos Preview. On the happy side, in its current form, Myth is nowhere near as scary as Tom Fridman (who worries about schoolchildren accidentally taking down power grids)