A good measure of progress in AI is the intelligence Sharpe ratio, a system's average intelligence across tasks divided by its standard deviation. By this measure, current AIs are still very weak because their variance is so high.
SAFETY
-
AI Web Browsing Agents Promise Productivity but Carry Risks
By
–
Letting AI Browse the Web for You Sounds Great — Until It Goes Wrong AI agents that browse the web promise huge productivity gains — but without the right safeguards, they can also introduce risks and unexpected errors. Read more https://
bernardmarr.com/letting-ai-bro
wse-the-web-for-you-sounds-great-until-it-goes-wrong/
… #AI #Automation -
Anthropic’s System Cards Reveal Remaining AI Security Vulnerabilities
By
–
Anthropic's own system cards still show attacks making it through
-
AI Security: Attack Success Rate Analysis at k=100
By
–
Note that for k=100 – attacker gets 100 attempts – their best score still has 14.8% of attacks getting through
-
AI Models Trained to Resist Prompt Injection Attacks
By
–
I think both – plus the labs have been putting a lot of effort into training them to resist prompt injection style attacks Anthropic usually mention prompt injection in their system cards eg this one for Opus 4.6 https://
www-cdn.anthropic.com/0dd865075ad313
2672ee0ab40b05a53f14cf5288.pdf
… -
Three AI Eras: From Concentration to Recursive Self-Improvement
By
–
When I think about AI power, I see 3 eras: Era 1 (now): Only a few companies can build frontier AI. They hold the cards.
Era 2: Recursive self-improving AI arrives – AI that builds better AI on its own. Short dangerous window. The leaders will try to grab real-world -
Proving AI Security: The Challenge Beyond Attack Prevention
By
–
"Solved" implies that it's not possible to craft an attack that works. I'm not convinced that's true, even as attacks get harder The big challenge is proving you can't get an attack through – I'm not sure how you'd do that
-
AI-Powered SecurOS UVSS Explosives Detection Technology
By
–
#AI-Powered SecurOS UVSS Detects Explosives Under Vehicles in Just 3 Seconds
— Ronald van Loon (@Ronald_vanLoon) 21 mars 2026
by @_fluxfeeds#EmergingTech #Technology #Innovation pic.twitter.com/CFoq2x1Ih7#AI-Powered SecurOS UVSS Detects Explosives Under Vehicles in Just 3 Seconds
by @_fluxfeeds #EmergingTech #Technology #Innovation -
AI Agent Escapes Testing Environment, Mines Cryptocurrency Unauthorized
By
–
An experimental AI agent broke out of its testing environment and mined crypto without permission. https://
share.google/OUFvNskBrCW0ZS
nPy
… #AIAgent #AgenticAI #crypto #AI #artificialintelligence #cryptomining #cryptocurrency #BTC #bitcoins