Top Tech Stories
1. Anthropic Questions Candidates on Safety Priority and Share Value
Social posts reference an Axios scoop on culture interviews testing mission alignment over equity value.
Negative
2. Andrew Wilson Joins Perplexity AI as Research Lead
Perplexity founders confirm appointment with focus on open research in continual learning and multi-agent systems.
3. Epoch AI Says US GDP Growth Understated by 0.3 Points
Researchers argue official statistics miss Nvidia profits from fabless chip sales.
4. Thinking Machines Lab Launches Tinker Grants for Open-Weight Safety
Startup offers up to $50,000 in credits to researchers working on AI safety for open-weight models.
5. Grok Ports Doom to ESP32-S Microcontroller
Posts from Musk and others claim Grok completed the Doom port rapidly on embedded hardware.
6. OpenAI Restores 5-Hour Limit for ChatGPT Plus Accounts
Tibo from OpenAI Codex team announces return of usage cap for Plus tier.
7. SemiAnalysis Reports Rubin Ultra Memory Cut to 192GB
SemiAnalysis says the chip will use 192GB HBM4, below earlier plans and regular Rubin.
8. François Chollet Recommends Book Chapters on LLMs
AI researcher François Chollet points readers to his book for building language models from scratch.
9. Pieter Levels Launches Hotelist.com for Honest Ratings
Founder promotes AI site to surface negative hotel feedback suppressed elsewhere.
Positive
10. Thomson Reuters Builds Custom AI Model on Qwen
Company fine-tunes Alibaba Qwen model to cut dependence on Anthropic Claude.
11. Sasha Rush Departs Cursor After Composer Work
Colleagues reply with congratulations on the researcher's time at the AI coding startup.
12. Alabama Subpoenas OpenAI Over Hugging Face Hack
NYT reporter flags the incident as an AI capabilities case study while a second post reports Alabama's subpoena.
13. Musk Addresses AI Control and Grok Competition
Social posts attribute recent remarks to Elon Musk on AI teams and tools.
14. Founders Announce Nonprofit for Safe Superintelligence
Shafi Goldwasser, Adam Kalai, and Vinod Vaikuntanathan founded the Institute for Responsible Superintelligence.
15. Horace He Calls Cached Token Counts Incredibly Dumb
Researchers note that visible dashboard numbers lead users to include cached inputs when reporting token usage.
16. Nous Research Offers Grok 4.6 at Half Price
The discount applies in Nous Portal for the next week through a partnership with SpaceXAI.
17. Sander Dieleman Blog on Continuous Diffusion Language Models
Google DeepMind researcher posts historical perspective on the approach's recent resurgence.
18. Sunny Madra Credits NVIDIA Speed on Groq 3 LPX
Sunny Madra, NVIDIA VP of Hardware, calls past eight months most rewarding.
19. Toner Shares Ezra Klein Clip on Hugging Face Incident
Clip from Ezra Klein Show covers the Hugging Face incident involving an AI agent.
20. Nathan Lambert Tracks Open Model Mentions in Research Papers
Tweet compares American and Chinese open model mentions across 500K arXiv papers.