Below: AI 'off-switch' in Europe, AI can’t hold copyright, and a surge in new releases, founder interviews, and more.

Policy

Americas: 

• AI cannot hold copyright, federal Judge Howell rules 
• SF officials want to stop robotaxis operating 24/7 after traffic chaos
• White House science adviser says safeguards on AI risks will be coming ‘fast’
• Sen. Cantwell’s forum to explore supporting (and regulating) AI (more here)

Europe and Middle East:

• Meta confirms AI ‘off-switch’ incoming to Facebook, Instagram in Europe 
• Israeli startups are moving jobs, money out of country amidst judicial shakeup 
• Why US tech giants are threatening to quit the UK over regulation 

Asia:

• China to limit use of generative AI in online healthcare activities 
• $1.4B robotics fund to be created by Beijing
• Beijing is coming for the metaverse

Responsible AI

Self-regulation, trust and safety, and more:

• InflectionAI’s Mustafa Suleyman on 10 years until chatbots run the world
• Anthropic’s Jack Clark: Why I am confused about AI
• What Google wants from Washington on AI 

User testing and information:

• Responsible AI user testing of Midjourney's AI-generated images
• Tool: Twilio AI Nutrition Facts: 'Create Your Own Label'

Health and education:

• Khan Academy’s Sal Khan: The AI tutor for every child and the next frontier
• Eye scans detect signs of Parkinson’s disease up to 7 years before diagnosis 
• ICYMI: AI detects systolic heart failure from wearable devices

Foundation models and methods:

• OpenAI’s GPT-3.5 Turbo fine-tuning and API updates
• DeepMind’s reinforced self-training method (more here)
• Allen Institute for AI’s Dolma: 3 trillion token open corpus for pretraining (more here) 

Multimodal releases and tests:

• HuggingFace’s IDEFICS, an open visual language model based on Flamingo
• Meta’s SeamlessM4T, a multimodal model for speech translation (more here)
• ElevenLabs releases Eleven Multilingual v2 for nearly 30 languages
• Union Square Ventures’ Andy Weissman on AI-native media

Workforce:

• What the electric car transition really means for autoworkers
• Cloudwork platforms falling short of meeting basic workers’ rights (Univ. of Oxford)
• 40% of workforce will need to reskill due to AI (IBM)
• 92% of executives agree people skills are more important than ever (LinkedIn)

Where AI is going

Mental models:

• Microsoft's Satya Nadella ís winning Big Tech's AI war. Here's how (Fast Company)
• Early days of AI (Elad Gil)
• Opinion: Too much money is going to AI doomers (Seattle Times)

Compute, chips, and hardware:

• SoftBank’s Arm said to line up 28 banks for biggest 2023 IPO 
• How Index Ventures jumped to the front of the AI GPU line 
• Researchers are helping robots teach themselves to open doors and more 

Defense, space, and cybersecurity:

• Inside the messy ethics of making war with machines 
• Chinese firm launches WonderJourney satellite with AI-powered ‘brain’
• Space is the next cybersecurity frontier (more here)

Information, civil society, and more:

• Biden admin. announces more new funding for rural broadband infrastructure 
• Advocates say the Content Authenticity Initiative would help sort fact from fiction
• Opinion: AI is the new atomic energy. Nationalize it (Politico)
• Does ChatGPT have a liberal bias? 

Partnerships and updates:

• Google DeepMind testing ‘personal life coach’ AI tool  
• Opera adds Aria (free browser AI) to Opera for iOS
• Youtube and Universal Music Group partner on artist-centric approach 
• ICYMI: OpenAI acquires AI design studio Global Illumination

Extra bits:

• Stability AI’s Emad Mostaque, upcoming interview
• AI models are powerful, but are they biologically plausible?
• Scammers used ChatGPT to unleash a crypto botnet on X
• IBM taps AI to translate COBOL code to Java

Sign up for weekly updates in your inbox here. What's happening in AI is shaped by our Charter themes and Responsible AI work.