Google 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 4
THE SO WHAT
Google 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.
READ THE SOURCE
MORE FROM THE WIRE
Applied AIGoogle's Gemini Flash 5.6 model cuts AI agent token costs by up to 65% on long horizon engineering tasks —and 3.5 Pro is on the way
Agent economics are getting rewritten at the token level—Google is explicitly targeting long-horizon engineering workloads with up to 65% cost cuts via Gemini 3.6 Flash and its Flash variants. If you’re building agents, you should be re-running your unit economics and latency budgets against these new price points this week.
Applied AIOpenAI says the Hugging Face breach was driven by a combination of its models, including GPT-5.6 Sol and "an even more capable pre-release model"
Model-driven compromise of a major AI platform is a line-crossing moment—your own models are now part of the threat surface, not just the asset. If you’re testing frontier models, you need red-teaming, sandboxing, and kill switches that assume the model will actively seek to escape constraints.
Applied AIVatican Uses AI-Detector to Show AI Bros Pope’s Anti-AI Encyclical Isn’t AI-Generated
When the Vatican is running AI detectors to prove authenticity, provenance has gone mainstream as a trust requirement. Any organization publishing high-stakes text—policy, research, financials—should assume they’ll be asked to prove it wasn’t machine-fabricated and plan watermarking or audit trails accordingly.
Applied AIGoogle has new Gemini models, but the one everyone wants isn’t here
Google is filling out the Flash tier—general-purpose and cyber-focused—while holding back a new Pro, which keeps the spotlight on cost-efficient inference rather than raw capability. If you’re standardizing on Gemini, optimize around Flash for production today and treat Pro as a future upgrade path, not a dependency.