Week 6: What's Happening in AI
New Executive Order, GPTBot, US AI laws, democracy reimagined, bionics and more -- covering policy, responsible AI, and where AI is going
Thanks for subscribing to the Responsible Innovation Labs Newsletter. We thought you may be interested in our new weekly series – check it out.
What’s Happening in AI: Covering the latest in policy, responsible AI, and where AI is going for the Responsible Innovation Labs community, every week.
Below: Executive Order on ‘national security technologies and products in countries of concern,’ state of US AI laws, bionic engineering, democracy reimagined, and more. First, this week’s big story: GPTBot: OpenAI releases new web crawler
Why this matters: Current web data can now be used to improve ChatGPT models. Previously, data ended in 2021. This is a first-of-its-kind jump towards being more competitive with search. Companies can use the Robots Exclusion Protocol to block the web crawler.
Consider: Should my company block ChatGPT’s access? How might AI-generated data affect model performance at scale? Can it be filtered out?
Documentation: GPTBot - OpenAI API
Policy
Americas:
The state of state AI laws (more here)
SEC’s Gensler on AI (more here)
Europe and the Middle East:
TikTok on fulfilling commitments under the EU’s Digital Services Act
Iran proposes AI surveillance and more in harsh new hijab law
ICYMI: Do foundation model providers comply with the draft EU AI Act?
Asia:
Explained: India’s new Digital Personal Data Protection Bill
ICYMI: China’s Interim Measures for the Management of Generative AI Services set to take effect Aug. 15
International organizations and more:
Plus: The White House just announced a new Executive Order to guide rule-making on US investment into ‘technologies and products (e.g., semiconductors and microelectronics, quantum information technologies, and AI) in countries of concern’ (China, Hong Kong SAR, Macau SAR).
Responsible AI
Security and safety:
The sound your keystrokes make is enough for AI to steal them (more here)
A new attack impacts AI chatbots—and no one knows how to stop it
Health and biomedical:
Data, language, and identity:
Clibrain introduces LINCE, the first LLM optimized for Spanish (more here)
8 months pregnant, arrested after false facial recognition match (more here)
Climate:
Sweetspot's co-founders on an AI search engine for the US government contract maze
Plus: Inversion mindset, explained
Where AI is going
Information, democracy, and open societies:
Interview with Google Jigsaw's Yasmin Green
Sam Altman is already nervous about what AI might do in elections
Deepfake speech is getting too good, fooling a quarter of listeners in study
Defense, transportation, and space:
Economy and workforce:
Will AI be an economic blessing or curse? History offers clues
Actors and writers aren't the only ones worried about AI, says new polI
Labor unions in the US and the Australian Council of Trade Unions move to protect workers
ICYMI: On the future of work in America and US labor shortage
Compute and chips:
Interview with AWS' Adam Selipsky (more here)
LK-99 was an online sensation — but replication efforts fall short
Plus: Twilio’s OpenAI partnership, Salesforce’s Einstein Studio, Huawei’s Voice Assistant Cecilia, Alibaba’s open-source models, Apple’s comments, Amazon’s comments
Thanks for reading. Feel free to get in touch with your questions, thoughts, or recommendations for next week.
If you've had this newsletter forwarded to you, you can sign up for the weekly edition here and read the Responsible Innovation Labs Charter for more about what shapes our Responsible AI work and this content.