Skip to content
Source · Daily Brief

AI Daily Brief — 15 May 2026

Gemini 3.5 launched — Google DeepMind framed it as “frontier intelligence with action”. Microsoft Research published further notes on AI delegation and long-horizon reliability. The Batch covered China thwarting Meta”s agentic ambition, US evaluations of upcoming models, and AI mammogram diagnoses.

Top stories

  • Gemini 3.5: frontier intelligence with action. Google DeepMind”s flagship launch — the action-oriented framing positions it head-on against agentic competitors. via Google DeepMind
  • Microsoft Research: Further notes on AI delegation and long-horizon reliability. Continued investigation into when humans should delegate to agents. via Microsoft Research
  • The Batch: China thwarts Meta”s agentic ambition; US evaluates upcoming models; AI mammogram diagnoses. Andrew Ng”s weekly survey. via The Batch
  • Dwarkesh Patel: Eric Jang on building AlphaGo from scratch. Long-form podcast on the practical engineering. via Dwarkesh Patel
  • Together AI + Pearl Research Labs partner to reduce AI inference cost. Joint announcement on cost reduction. via Together AI

Who shipped

Google DeepMind dominated with the Gemini 3.5 launch. xAI connected Grok to the Hermes Agent. Amazon Science published on making LLMs faster without sacrificing accuracy.

Quiet corners

OpenAI, Anthropic, Mistral, Meta (Llama), DeepSeek were absent — yielding the day to Gemini 3.5.

By the numbers

  • 9 stories tracked across 7+ sources
  • Most-mentioned model: Gemini 3.5
  • Most-mentioned lab: Google DeepMind
  • Notable absences: OpenAI, Anthropic, Mistral, Meta, DeepSeek

Compiled by AI Feed”s editor from all 9 headlines published on 15 May 2026.