Can LLMs work together well? New resources from YC, Anthropic. VCs and experts shape responsible AI guidelines for startups. Founders on synthetic biology, national security, human relationships, more. EU regulation and compliance. AI’s carbon footprint.

Policy

Americas

• Responsible Innovation Labs co-founder Hemant Taneja and U.S. Department of Commerce Secretary Gina Raimondo describe collaboration on responsible AI with a cross-sector coalition of VCs and experts at Fortune’s CEO Initiative (more here and here). RIL will be sharing more about this work with our community in the coming weeks.
• Microsoft CEO testifies on limits of AI in Google’s antitrust trial
• Conversations with Senators Young, Markey, and Warner at POLITICO’s AI & Tech Summit

Europe

• UK to pitch new ‘AI Safety Institute’ to allies at upcoming AI Summit (more here)
• Google agrees to reform its data terms after German antitrust intervention
• TikTok tests an ad-free subscription tier and Meta plans similarly in response to EU’s DSA
• Uber still dragging its feet on algorithmic transparency in GDPR, Dutch court finds

More

• Millions of sellers stranded as TikTok Shop shutters in Indonesia

Responsible AI

Safety and alignment

• NIST, DeepMind General Counsel, Rep. Lieu re: can you believe in machines?
• Google is opening up its generative AI search experience to teenagers

Sciences and health

• When AI and genomics collide with Insitro founder Daphne Koller
• Scaling synthetic biology with Ginkgo Bioworks co-founder Jason Kelly

Information

• How to stop AI deepfakes from sinking society — and science
• There's an election 'megacycle' coming. The world isn't ready (more here)

Climate

• AI's carbon footprint sparks debate
• AI is getting better at hurricane forecasting

Security and defense

• Conversation with Anduril founder Palmer Luckey (and news on their raise)
• US DOJ urged to investigate gunshot detector purchases

Identity and data

• AI bias tests gloss over a crucial aspect of skin color, Sony research claims
• Medium says default no to AI training on your stories
• Meta said it's been using your public Instagram photos and Facebook posts to train its AI

Resources

• GenAI founder matching meet-up (Y Combinator)
• Challenges in evaluating AI systems prompt (Anthropic)
• Meeting the FedRAMP FIPS 140–2 requirement on AWS (Medium)
• Engineering for Claude’s long context window (Anthropic)

Where AI is going

Capabilities

• ‘Can LLM agents work together well?’, collaboration mechanisms explored in new paper
• Bumble founder Whitney Wolfe Herd on AI in matchmaking at Code 2023
• Why the future of video is synthetic with Synthesia co-founder Victor Riparbelli
• SoftBank CEO Masayoshi Son says AGI will come within 10 years
• Can AI turn every employee into a developer?

Autonomous and smart

• Google DeepMind on scaling up learning across many different robot types (more here)
• 'The Android of agriculture': The future of farming with Monarch Tractor founder Praveen Penmetsa
• Scaling GAIA-1: 9-billion parameter generative world model for autonomous driving

Economics and competition

• Brookings says without intervention, GenAI jobs will continue to cluster in the big tech hubs
• GPU shortage, AI regulation, and Nvidia competition with AMD CEO Lisa Su
• Bing's quest to beat Google and the future of AI art with Microsoft CTO Kevin Scott

Releases and raises

• Releases: Perplexity’s pplx-api, Canva’s Magic Studio, Zoom Docs, Meta’s Quest 3 and advertising features, Reka’s Multimodal AI Assistant, LinkedIn’s Career Coach
• Raises: Gradient ($10M, DevOps), Lemurian Labs ($9M, chips), Vera ($2.7M, AI governance)‍
• OpenAI and Jony Ive in talks to raise $1B from SoftBank for ‘iPhone of AI’‍
• AI-powered Metropolis raises $1.7B to acquire SP Plus‍
• RunwayML partners with Canva