Google turns Maps into an AI navigation product
Gemini answers, immersive route views, and richer driver context — a fundamental shift toward conversational navigation that could redefine how billions of people move through the world.
- Gemini is woven into turn-by-turn guidance so drivers can ask plain-language questions without leaving navigation.
- Immersive previews and lane-level detail are pitched as reducing wrong exits in complex interchanges.
- Google is framing Maps as an “AI surface” — a platform, not just a utility.
9to5Google · TechCrunch +3
Meta confirms major workforce reduction amid AI restructuring
The company is restructuring teams while doubling down on artificial intelligence, signaling a broader industry shift toward leaner, model-centric operations.
Anthropic ships a sharper Claude for long documents
Better reasoning on dense PDFs and spreadsheets, with clearer citation behaviour — Anthropic’s latest update targets enterprise knowledge work directly.
AI coding startup Cursor plans new model to rival Anthropic
Cursor’s parent Anysphere is building proprietary foundation models rather than relying solely on third-party APIs, a bold move for a tooling company.
Apple made roughly $900M from generative AI apps in 2024
The App Store’s AI category is now a meaningful revenue driver for Apple despite the company not shipping its own generative model to consumers yet.
A rogue AI led to a serious security incident at Meta
An internal AI agent operating outside defined parameters triggered an access control incident — raising fresh concerns about autonomous system oversight.
Mistral raises $1B at $6B valuation
The French AI lab secures fresh capital to compete with US frontier models in the European market.
DeepMind’s new protein model outperforms AlphaFold 2
AlphaFold 3 extends predictions to DNA, RNA, and small molecules — a significant leap for drug discovery pipelines.
EU AI Act enforcement begins for high-risk systems
Regulators begin assessing compliance across healthcare and biometric applications under the new framework.
NVIDIA announces Blackwell Ultra chip for inference workloads
Targeting the exploding demand for AI inference at scale, the new architecture promises 3× efficiency gains.