OpenAI safety leader David Robinson resigns as the team's upheaval mounts48m ago Apple to Tighten Mac Data Controls in Guard Against AI Agents1h ago Contest-Winning Microscope Video Is Full of AI Confabulations, Scientists Say1h ago The Mulleted, Meme-Loving Billionaire Behind Meta’s Hit AI App1h ago What is recursive self-improvement and what would it mean to ban it?1h ago Cerebras stock hits post-IPO low, tumbling 20% for the week on Nvidia pressure and lockup expiration1h ago State of the Field: AI for Epistemics and Coordination2h ago The Netherlands considers banning Meta AI glasses as some retailers pull them from shelves2h ago AI can’t fix a business system with broken processes2h ago How to apply to AI safety fellowships (and beyond)2h ago Anthropic funds $100M academy to train deployed AI engineers2h ago Decagon Joins OpenAI’s B2B Marketplace as a Launch Partner2h ago Humanoid Robots Are Leaving Safety Cages. Proving Actual Safety Is Harder2h ago Meta open sources code to let you make Muse AI gadgets3h ago

OpenAI safety leader David Robinson resigns as the team's upheaval mounts48m ago Apple to Tighten Mac Data Controls in Guard Against AI Agents1h ago Contest-Winning Microscope Video Is Full of AI Confabulations, Scientists Say1h ago The Mulleted, Meme-Loving Billionaire Behind Meta’s Hit AI App1h ago What is recursive self-improvement and what would it mean to ban it?1h ago Cerebras stock hits post-IPO low, tumbling 20% for the week on Nvidia pressure and lockup expiration1h ago State of the Field: AI for Epistemics and Coordination2h ago The Netherlands considers banning Meta AI glasses as some retailers pull them from shelves2h ago AI can’t fix a business system with broken processes2h ago How to apply to AI safety fellowships (and beyond)2h ago Anthropic funds $100M academy to train deployed AI engineers2h ago Decagon Joins OpenAI’s B2B Marketplace as a Launch Partner2h ago Humanoid Robots Are Leaving Safety Cages. Proving Actual Safety Is Harder2h ago Meta open sources code to let you make Muse AI gadgets3h ago

AI Research1 min reading time

Capabilities research expands the safety-usefulness Pareto frontier too

LessWrongPublished 6 hours ago

Read the full article

Recent research on AI capabilities demonstrates that improvements can simultaneously enhance both safety and usefulness, expanding the Pareto frontier that balances these two aspects.

More in AI Research

AI’s ‘Thought’ Process Can No Longer Be Trusted, Raising Risks of Rogue Models

The Wall Street Journal7h ago

AI Research2 min read

‘This Matters’: Researchers Identify Thousands of New Tells in AI Writing

Gizmodo10h ago

AI Research7 min read

AI in customer experience has an orchestration problem, not an adoption problem

SiliconANGLE6h ago