Mercedes-Benz rolls out conversational models that run directly on its cars instead of the cloud through a partnership with MIT spinout Liquid AI
THE SO WHAT
Mercedes running conversational models directly in-car is a clear proof point that on-device LMs are ready for production-scale, safety-relevant environments. If you build for automotive or other constrained hardware, start a serious evaluation of SLMs and edge inference instead of assuming perpetual cloud dependence.
READ THE SOURCE
MORE FROM THE WIRE
Applied AIPartnering with Edelbyte: Data Infrastructure That Keeps Up With AI's Curiosity
If agents are going to hammer your internal systems with orders of magnitude more queries than humans ever did, the real constraint becomes data infra cost and latency, not model quality. Teams building agents should be budgeting for a new class of OLAP-style infra like Edelbyte’s, not assuming the existing warehouse can take the load.
Applied AIOpenAI’s GPT-6 Cyber found 81 new Firefox flaws and is rated ‘Critical’
GPT-6 Cyber surfacing 81 new Firefox vulnerabilities is a concrete proof point that offense-grade AI is now a security product category, not a demo. CISOs need to assume attackers will have comparable tools and start budgeting for AI-native red teaming and continuous code scanning this quarter.
Applied AISiri AI's Biggest Upgrade Yet + John Ternus's First Apple Earnings Call
Siri gaining real app control and multi-step task execution in iOS 27.3 moves Apple from voice commands to on-device agentic behavior. If you’re an iOS-first product, your integration surface is about to shift from UI taps to intent APIs and background workflows.
Applied AIGemini Omni 1.2 Flash: Fast and cost-effective with longer clips
A faster, cheaper Gemini Omni 1.2 Flash with longer context pushes multimodal reasoning closer to real-time, high-volume workloads. If you’re cost-gated on video or long-document use cases, it’s time to re-run your unit economics and latency benchmarks against this tier.