0
On the Wire now
Free
Applied AIGoogle launches Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber, and says it has started its "most ambitious pre-training run yet" for Gemini 4Google is pushing a barbell strategy—cheap, fast Flash variants for mass deployment now, while quietly spinning up a huge Gemini 4 pre-train for the next capability jump. If you’re building on Gemini, optimize around Flash economics this year but keep your high-stakes roadmap modular enough to swap in Gemini 4 without a full rewrite.9to5GoogleStartups & VentureA startup building a map to help you make plans with friends has doubled its fundingCorner’s $7.5M for a social map of saved and shared spots is another bet that location is the next social graph primitive. If you run local or experiential products, assume your discovery and attribution will increasingly run through map-native feeds rather than generic social timelines.Business InsiderApplied AIUniversity of Tennessee Research Foundation sues Anthropic in Delaware for allegedly infringing neural network patents, the first such case against AnthropicA university foundation bringing neural network patent claims against a major lab is a warning that core AI IP is moving into active litigation, not just licensing. If you’re training or serving models at scale, you need an IP map — what’s open, what’s licensed, and where you’re exposed if academic or corporate holders start enforcing.ReutersApplied AITreasury Secretary says its 'unacceptable' that Chinese open source AI models may have been built on 'IP theft'Calling alleged IP theft in Chinese open-source AI models “unacceptable” at Treasury level is a precursor to concrete tools — sanctions, export controls, or procurement bans. Any team consuming foreign open models needs to document provenance and be ready with substitutes if specific ecosystems get swept into policy action.Business InsiderApplied AIThe AI Copyright Lawsuits Have Finally Produced an Actual Payout$1.5B moving from an AI lab to rights holders turns copyright risk from theoretical to priced-in. If your models or data vendors touch copyrighted corpora, you now have a reference number for downside exposure — legal, PR, and cash — and boards will start asking for that line item.Gizmodo AIApplied AIGoogle releases three new Gemini models — but no 3.5 ProGoogle doubling down on 3.6 Flash, 3.5 Flash-Lite, and Flash Cyber — and still holding back 3.5 Pro — says the near-term fight is on cost, latency, and specialization, not max IQ. If you’re building on Gemini, design for a portfolio of cheap, task-tuned models rather than betting your roadmap on a single flagship that keeps slipping.TechCrunch AI

Latest Weekly Signal & Field Reports