Frontier Daily (Nov 12): 9 items from Simon Willison, DeepMind, Hacker News and more

Every day Laojin pulls the official channels and compiles the past 48 hours of AI news into a short brief: curation only, no commentary, with the original link and publish time attached to every item.

1. Anthropic's IPO roadshow video

Anthropic's IPO roadshow video reveals previously unreleased details on its multi-chip compute strategy, including training and serving Claude across AWS Trainium, Google TPUs, NVIDIA GPUs and AMD GPUs.

Source: Simon Willison's Weblog · 2026-11-11 · Original

2. AlphaGenome 2: Deciphering how cells read DNA to better understand disease

Google DeepMind introduces AlphaGenome 2, which aims to predict how DNA sequences and their variations affect specific human cell types and disease.

Source: Google DeepMind Blog · 2026-11-11 · Original

3. Atuin Threads: All your agent sessions on every machine

Atuin launches Threads, a tool for keeping all your agent sessions available on every machine.

Source: Hacker News · 2026-11-11 · Original

4. Google SWE points 48 AI agents at NASA Lidar, finds unmapped Maya pyramid in 72h

A report describes Google SWE directing 48 AI agents at NASA Lidar data and finding an unmapped Maya pyramid within 72 hours.

Source: Hacker News · 2026-11-11 · Original

5. ChatGPT Plus for veterans and servicemembers

OpenAI highlights its ChatGPT Plus offering for veterans and servicemembers, with more than 20,000 having claimed a free year to prepare for civilian careers.

Source: OpenAI Blog · 2026-11-11 · Original

6. llm-gemini 0.36

llm-gemini 0.36 ships as Gemini 4 Carbon becomes generally available at $2/$10 per million input/output tokens, scheduled to double to $4/$20 on December 1st.

Source: Simon Willison's Weblog · 2026-11-10 · Original

7. Introducing Businesses in ChatGPT

OpenAI introduces Businesses in ChatGPT for managing business information, viewing how a business appears, and turning interest into action.

Source: OpenAI Blog · 2026-11-10 · Original

8. Gemini Robotics 2.5 accelerates real-world general dexterity

Google DeepMind releases Gemini Robotics 2.5, described as its most capable vision-language-action model for robot dexterity, reliability and safety.

Source: Google DeepMind Blog · 2026-11-09 · Original

9. Fable 5.2 and Jagged AGI

The author reviews Claude Fable 5.2 ahead of launch and argues the model represents a 'jagged AGI' that still trails humans in important areas under a computer-based definition of AGI.

Source: One Useful Thing · 2026-11-09 · Original

---

_Sources: official blogs of OpenAI, Google DeepMind and Qwen; official RSS of Simon Willison, Latent Space, Import AI and One Useful Thing; Anthropic site monitoring, top AI discussions on Hacker News and open-source updates from key AI orgs on GitHub. Mistral / Meta AI have no stable public feed; X sources are curated by hand. This is a source digest — no commentary from Laojin._