Week 10: What’s Happening in AI

Regulators back in session, AI alignment, molecules-to-aromas, LLMOps, new releases, and more.

Welcome to this week’s edition of What's Happening in AI, covering the latest in policy, responsible AI, and where AI is going for the Responsible Innovation Labs community.

Below: Regulators back in session, AI alignment, molecules-to-aromas, LLMOps, new releases, and more

Policy

Americas: 

• Senate Committee to hold AI hearing on September 12

• Governor Newsom signs Executive Order on AI in California (more here)

• Senator Cassidy releases paper on AI in health, education, labor (more here)

• Prosecutors in every state push to combat AI child exploitation

• ICYMI: Musk and Zuckerberg are among the tech execs who will attend Senator Schumer’s first AI forum on September 13

Europe:

• UK government sets out AI Safety Summit ambitions 

• 5 things to know as the EU Digital Markets Act ramps up (more here)

• Report: UK Members of Parliament identify 12 challenges of AI governance

• Spain launches AI regulation agency in bid to become industry leader

• Britain admits defeat in controversial Online Safety Bill fight 

Asia:

• Baidu CEO says more than 70 LLMs approved and released in China

• NVIDIA CEO meets with India Prime Minister Narendra Modi 

Responsible AI

Governance and alignment:

• AI will scramble geopolitical power if left unchecked, Inflection AI’s CEO says

• OpenAI’s moonshot: Solving the AI alignment problem

• Survey: Experts favor new U.S. agency to govern AI

Health and science:

• This neural net maps molecules to aromas

• Could AI be the spark that ends the opioid epidemic?

• Sizing up AI's promise and limitations in cancer detection

Climate:

• New AI-generated map of the world’s trees and renewable energy projects

Resources:

• FACET: Benchmarking fairness of vision models (Meta)

• How to foster belonging and inclusion, with Ellen Pao (Multitudes)

• A list of open LLMs available for commercial use (Github Repo)

• Belebele: New benchmark for evaluation of massive, multilingual natural language understanding (Meta) 

• Teaching with AI (OpenAI)

Where AI is going

Society and politics:

• The who, where, and how of regulating AI – and where is the global south?

• Why the West is concerned about the UN cybercrime treaty

• Meet Aleph Alpha, Europe’s answer to OpenAI

• AI and the new digital Cold War

Defense:

• How autonomous drones beat the best human racers

• AI brings the robot wingman to aerial combat

Information and communication:

• New photon polarizer lights way to quantum communications

• The scoop on Gen Z and how they are rewriting the rules of the internet

• US Copyright Office’s public comment on AI and copyright opened August 30

Releases and partnerships:

• HuggingFace’s Training Cluster as a service

• Google’s SayTap: Language to quadrupedal locomotion

• Refact LLM (1.6B code model) reaches 32% HumanEval

• TinyLlama 1.1B parameter project kicked off September 1

• Falcon 180B introduced by UAE's Technology Innovation Institute

• Baidu launches Ernie chatbot after Chinese government approval 

• Tencent’s Hunyuan foundation AI model

• Alibaba Cloud’s Qwen-LM/Qwen-VL

• Salesforce revs its VC engine in the AI race

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