Groq Raises $650M After Nvidia's $20B Deal, Coders Refuse to Work Without AI, Apple Brings Gemini to iPhone — AI News Briefing

Top 7 Stories

1. After Nvidia’s $20B Deal, AI Chip Startup Groq Raises $650M to Pivot to AI Inference

AI chip startup Groq is reportedly raising $650 million in internal funding as it pivots from hardware manufacturing to focus on AI inference — the process of running trained models to generate responses. The move follows Nvidia’s staggering $20 billion “not-acqui-hire” deal, which reshaped expectations for AI chip valuations and competitive dynamics.

Groq’s pivot reflects a broader industry shift as companies race to address the inference bottleneck that grows more acute with each new model release. While training gets most attention, running models at scale for millions of users requires fundamentally different chip architectures. Groq’s approach emphasizes low-latency inference, positioning itself as a complement rather than competitor to Nvidia’s training-focused GPUs.

Sources: TechCrunch, Axios

2. Coders Are Refusing to Work Without AI — and That Could Backfire

A growing number of software developers are refusing to work without AI coding assistants, but researchers warn this trend may be producing faster code at the expense of quality. The phenomenon, dubbed “tokenmaxxing” by some observers, sees developers measuring productivity by volume of AI-generated output rather than correctness, maintainability, or architectural soundness.

The concern is that AI-generated code often works in the short term but introduces subtle bugs, security vulnerabilities, or technical debt that compounds over time. As AI coding tools become entrenched in development workflows, teams may face a reckoning when the accumulated quality issues surface in production systems — potentially years after the code was written.

Sources: TechCrunch

3. Cognition’s Scott Wu Says AI Coding Agents Shouldn’t Replace Humans

Scott Wu, founder and CEO of Cognition Labs — maker of Devin, the first widely successful AI coding agent — has publicly stated that AI coding agents are not designed to supplant human programmers. Wu’s comments come amid growing industry debate about whether autonomous coding agents will ultimately displace software engineering roles.

Cognition’s positioning emphasizes augmentation over replacement, with Devin designed to handle repetitive coding tasks, debugging, and code review while leaving architectural decisions and complex problem-solving to human developers. Wu’s stance contrasts with some enterprise AI adopters who are actively pursuing headcount reduction through AI agent deployment.

Sources: TechCrunch

4. Box CEO Aaron Levie Warns of “AI Psychosis” as Companies Cut Staff for AI Agents

Box founder and CEO Aaron Levie has coined the term “AI psychosis” to describe companies making aggressive AI replacement decisions without understanding what their employees actually do. The comment follows ClickUp’s recent announcement that it cut 22% of its workforce in favor of AI agents, part of a broader 2026 trend where tech layoffs are already nearing all of 2025’s total.

Levie argues that executives championing AI-driven layoffs are often the least equipped to evaluate which roles can genuinely be automated. The warning resonates as AI agent deployments in enterprise settings reveal gaps between marketing promises and practical reality — agents that excel at demo tasks frequently struggle with the nuanced, context-dependent work that makes up most professional roles.

Sources: TechCrunch, Equity

5. Apple Reportedly Working to Cram Massive Gemini Model Into iPhone for New Siri

Apple is reportedly working with Google to integrate a compressed version of Google’s Gemini AI model into future iPhones, aiming to power a significantly upgraded Siri experience. The effort to run a large language model on-device represents a major engineering challenge, requiring novel model compression techniques to fit within iPhone memory and thermal constraints.

On-device AI processing would give Apple a privacy advantage over cloud-dependent competitors while enabling faster, more responsive voice interactions. The collaboration between Apple and Google — typically fierce rivals in mobile — underscores how AI capabilities are reshaping traditional competitive boundaries in the tech industry.

Sources: Ars Technica

6. Illinois Passes Landmark AI Regulation as Federal Control Weakens

Illinois has passed a landmark AI regulation law that significantly curtails federal control over AI policy, marking a turning point in the balance of power between state and federal governments on technology regulation. The legislation establishes state-level requirements for AI transparency, bias auditing, and consumer protection that go beyond current federal frameworks.

The move reflects a growing trend of states taking AI governance into their own hands as federal policy remains fragmented. Industry groups are watching closely as Illinois’s approach could become a model for other states, potentially creating a patchwork of AI regulations that companies must navigate — similar to how privacy law evolved after California’s CCPA.

Sources: Ars Technica

7. Google to Remake Search With Agentic AI, Adds More Sources to AI Overviews

Google is preparing a fundamental overhaul of Search powered by agentic AI capabilities, shifting from traditional keyword-based results to AI-driven task completion. Alongside this shift, Google announced it will link to more source websites in its AI Overviews, a response to ongoing criticism that the feature was diverting traffic away from content creators.

The agentic search vision envisions users asking complex, multi-step questions and receiving actionable results — booking travel, comparing products, summarizing research — rather than lists of links. For publishers, the increased source linking offers some relief, but the fundamental shift toward AI-mediated search access still threatens the traditional SEO-driven traffic model that many websites depend on.

Sources: Ars Technica

Trend Watch

What to Watch

• Groq’s inference roadmap: Whether the $650M raise translates into production-grade inference chips that can compete with Nvidia’s offerings. Watch for customer announcements and benchmark results.
• AI coding quality metrics: As more developers rely on AI, expect new tools and standards for measuring code quality, security, and maintainability of AI-generated output.
• Enterprise AI agent ROI: Companies like ClickUp betting big on AI agent replacement will face scrutiny over actual productivity gains vs. morale and quality costs.
• Apple’s on-device AI capabilities: If Apple successfully ships Gemini-powered Siri on-device, it could force Google and OpenAI to accelerate their own edge AI strategies.
• State AI regulation cascade: Illinois’s law may trigger similar legislation in California, New York, and Texas, creating compliance complexity for AI companies.
• Google Search traffic shifts: As agentic AI takes over more queries, watch for publisher revenue impacts and potential antitrust scrutiny of Google’s search monopoly.