Key updates include Anthropic's public benefit fund, Meta's terms change on web scraping, Replit switching to OpenAI as default model, Disney hiring most of Character.AI staff, and a new llm-tools-datasette release.
Anthropic announced it will allocate five percent of its equity to a fund aimed at supporting public benefit initiatives related to AI.
Understanding corporate governance moves helps gauge long-term AI safety commitments.
Source: TechCrunch AI
Meta updated its terms of service to delete clauses that had suggested the company could scrape the web for AI training data.
Policy changes affect how developers can legally use web data for model training.
Source: TechCrunch AI
Replit announced that OpenAI models will now be the default offering on its platform, though Anthropic models remain available as an option.
Platform defaults influence which APIs developers learn and integrate first.
Source: TechCrunch AI
Disney filed to employ most of Character.AI employees, indicating a strategic acquisition of talent in conversational AI.
Talent shifts signal where industry expertise and future product focus will concentrate.
Source: TechCrunch AI
The llm-tools-datasette plugin was updated to version 0.1.1, fixing compatibility with the newer httpx2 dependency in LLM 0.33.
Keeping tooling compatible ensures smooth integration of LLMs with SQLite-based data stores.
Source: Simon Willison
Study how to connect large language models to relational databases using the llm-tools-datasette plugin and explore the updated API surface. Then review the recent Meta terms change to understand legal constraints on web-scraped training data.
Build my AI Engineer plan