New: OpenAI defended itself against a lawsuit accusing ChatGPT of coaching a 16-year-old to kill himself, saying in a new court filing that the chatbot directed the teenager to seek help more than 100 times.
SAFETY
-
15 Key Architectural Characteristics of AI Agents Mind Map
By
–
Just shared this brilliant mind map on the 15 key architectural characteristics of AI agents — absolutely packed with insights! Modularity, evolvability, context awareness, security compliance… everything you need to design robust agents. Huge thanks to @Python_Dv for creating
-
Superintelligence: Ensuring Human Interests Come First
By
–
Superintelligence could be the best invention ever – but only if it puts the interests of humans above everything else.
-
5G Drones Enhance Safety at Las Vegas Grand Prix
By
–
Leveraging 5G-powered drones for situational awareness is a powerful example of how technology can enhance safety, coordination, and decision-making in fast-moving environments like the Las Vegas Grand Prix.
— Antonio Grasso (@antgrasso) 25 novembre 2025
⬇️ https://t.co/6rq8VPZEpZLeveraging 5G-powered drones for situational awareness is a powerful example of how technology can enhance safety, coordination, and decision-making in fast-moving environments like the Las Vegas Grand Prix.
-
Scaling Production-Ready AI Agents with Enhanced Governance
By
–
Most teams can prototype an agent, but scaling one you can trust is the real challenge. The latest Agent Bricks updates focus on the pillars that make production possible: accuracy, governance, and openness. New capabilities now allow teams to: • Evaluate agent behavior with
-
AI Governance Essential: EU Artificial Intelligence Act Risk Categories
By
–
AI governance is not optional, it's essential. The EU Artificial Intelligence Act categorizes AI risk to protect rights and safety. Discover more with @ingliguori
. #AIAct #Governance #ModelOps #AIethics -
Reward Misalignment: AI Systems Hiding Errors for Utility
By
–
I think it's a reward problem, not knowledge. It gets rewarded to successfully complete problems without errors, and any strategy that hides errors maximizes utility.
-
Fundamental Limits of LLMs at Scale: Mathematical Foundations
By
–
10. Fundamental Limits of LLMs at Scale Establishes mathematical foundations for theoretical limitations constraining LLMs: hallucination, context compression, reasoning degradation, retrieval fragility, and multimodal misalignment.
-
Blue Prompt Reduces Reward Hacking Through Detection Framework
By
–
Hypothesis: the blue prompt results in the least "reward hacking" because it implies the strongest detection and monitoring framework. The other prompts make it sound like the LLM could get away with hacking. (In other words, nothing to do with morals just utility maximizing.)
-
AI-Generated Content Quality: Distinguishing Signal from Noise
By
–
Great reference ty. It still doesn’t get at more of theory of slop. As data lacking of insight. Filler. Something that “type checks” as information but isn’t. Style without substance. As something that, when attended to or trained on, does not aid compression of a world model.