AI group petitions for Elon to lead AI policy in Trump administration

Today’s Highlights:

📰 News: AI advocacy group petitions for Elon to lead AI policy in Trump administration + Google DeepMind open-sources AlphaFold 3

💰 Funding: 11x raises $50Bn from a16z

⚡️ Top News Stories:

1. AI advocacy group, Americans for Responsible Innovation (ARI), petitions for Elon Musk to lead AI policy in the Trump administration, citing his tech influence.

• ARI argues Musk’s experience with OpenAI and AI safety advocacy positions him well to guide AI policy, despite his potential conflicts of interest as the head of AI firm xAI.

• The ARI petition acknowledges Musk’s conflicts but asserts they can be managed, suggesting Musk’s tech expertise and safety concerns could be pivotal for responsible AI regulation.

• ARI believes he might still support key AI safety agencies like NIST, potentially shaping U.S. AI policy without undermining regulatory bodies critical to AI safety.

• Musk’s call for a pause on advanced AI development and previous support for AI safety legislation signal an evolving but unclear stance on specific policies, leaving room for influence from public debate.

2. Following the Nobel Prize awarded to its creators, Demis Hassabis and John Jumper, Google DeepMind has open-sourced AlphaFold 3 for academic use, a diffusion-based, physics-aligned model that predicts complex molecular interactions, advancing drug discovery by reducing lab time and costs while improving accuracy over traditional methods.

3. The major AI companies are moving away from the "bigger is better" model due to scaling challenges and diminishing returns, instead focusing on “test-time compute” techniques that enhance model performance during inference—such as OpenAI’s o1 generating multiple solutions in real time—shifting hardware demand from Nvidia's training chips to distributed “inference clouds” and opening market competition.

4. Google has launched “Learn About,” an experimental AI tool based on the LearnLM model, designed to provide educationally-focused, interactive answers with visual aids, textbook-style context boxes, definitions, and relevant side topics for a richer learning experience.

5. X is expanding access to its Grok AI chatbot by offering a free version with usage limits—allowing 10 queries every two hours for Grok-2, 20 for Grok-2 mini, and three image-based questions daily—with availability in certain regions like New Zealand, requiring accounts to be seven days old and phone-number verified.

6. The Washington Post has introduced “Ask The Post AI,” a generative AI tool that provides users with concise, trustworthy answers and curated information by leveraging the publication’s extensive journalism database and commitment to fact-based reporting.

7. Google’s AI-powered flood forecasting model now covers 100 countries, reaching over 700M people with 7 day lead predictions—previously only possible for 5 day forecasts—and is enhanced by new data, model improvements, and “virtual gauges” in 150 countries, with forecast data accessible through a new API and the GRRR dataset for research on flood patterns.

8. The U.S. has directed TSMC to halt shipments of advanced AI and GPU chips (7 nm or smaller) to Chinese customers, including Huawei, starting this week, following TSMC's report of a chip in Huawei's AI processor that may violate U.S. export controls, with Huawei still on a restricted trade list.

9. MIT researchers developed LucidSim, an AI platform that uses physics simulations and ChatGPT-generated scenes to train four-legged robots in diverse simulated environments, achieving over 70% higher task success in real-world performance.

10. Ai-Da, the world's first humanoid robot artist, made history when her painting of Alan Turing, titled "A.I. God," sold at Sotheby's for over $1M, significantly surpassing its estimated price and marking a notable moment in the art world's relationship with AI. 

11. The Beatles' track “Now and Then,” nominated for Record of the Year and Best Rock Performance, marks the first Grammy nomination involving AI, where AI-assisted stem separation was used to isolate and clean up John Lennon’s 60-year-old demo vocals, not for deepfake purposes.

12. Lilian Weng, OpenAI’s VP of Research and Safety, has departed after seven years, citing a desire to "reset and explore something new," joining a wave of exits that reflect growing concerns over AI safety priorities.

13. During the recent election, ChatGPT directed around 2M users to trusted news sources like Reuters and the Associated Press instead of providing real-time updates, declined about 250K image requests to block deepfakes, while competitor Perplexity launched an election hub that garnered 4M views, highlighting rising trust in AI for election information.

💰 Top Funding News:

1. 11x, which uses AI to create fully autonomous "digital workers" rather than traditional software, aiming to replace human-driven tasks in sales and revenue generation, raised a $50M Series B, led by Andreessen Horowitz.

2. Amazon is in talks to significantly increase its investment in Anthropic, potentially committing billions more on top of last year’s $4Bn deal, with conditions that Anthropic use Amazon’s silicon on AWS rather than Nvidia chips; this new funding could support Anthropic’s $40Bn valuation goal and strengthen its competitive position, having raised $9.7Bn compared to OpenAI’s $21.9Bn.

3. Filigran, which uses AI and data engineering to enhance cybersecurity capabilities, including threat-driven risk assessment and security validation, raised a $35M Series B, led by Insight Partners.

4. UnifyApps, which uses AI to connect companies' SaaS applications and data, allowing them to build custom AI chatbots that reliably interface with enterprise information, raised a $20M Series A, led by ICONIQ Growth.

5. Cogna, which leverages gen-AI and natural language processing to create custom enterprise software for clients, allowing non-technical teams to specify needs in plain language, raised a $15M Series A, led by Notion Capital, w/ Hoxton Ventures and Chalfen Ventures.

That's all for today's email! If you want more please follow us at the social channels linked below, or check out our website!

Share our newsletter: If you like our work please share/forward this email with your friends, colleagues, and family. It's the best way to support us!

If this email was forwarded to you please sign up here to continue receiving them.

Want your content, product, jobs, or event featured in our newsletter? Reply to this email with the details, and our team will reach out to you.

Do you use AI for work? Tell us how, and you could be featured in our newsletter!

Check out our website for more resources, including a list of AI investors, products, events, and twitter follows.

For an archive of all our posts, click here.

We'd love to hear from you! You can always leave us comments or feedback by replying to this email!

Powered by AI. Curated and edited by Humans.