AI pendants, job losses, coding rot, and a $650M Groq bet
Meta's building a wearable, companies are firing people for AI agents, and coders can't remember how to code without a copilot.
Subscribe free →Meta's Next Hardware Bet Is a Pendant That Talks Back
Meta is reportedly developing an AI-powered pendant, doubling down on its wearable AI hardware strategy after the Ray-Ban smart glasses gained traction. Few details are public yet, but it signals Meta is serious about getting AI onto your body, not just your feed.
Why it matters: The race for ambient AI hardware is on — whoever owns the wearable wins the context layer, and Meta wants that to be them.
Google's Gemini Spark Is Useful — So Why Does It Exist?
Google's new 24/7 AI assistant Gemini Spark handles inbox summaries, schedules, and local event planning competently enough to earn a "actually pretty useful" verdict from TechCrunch's reviewer. The catch: it's a separate product from the existing Gemini assistant, and nobody can quite explain why Google carved it off.
Why it matters: Google's AI assistant strategy is becoming a confusing product maze, and fragmentation at this stage could hand the edge to more focused competitors like OpenAI's GPT-4o voice mode.
"AI Psychosis" Is Real — and It's Getting People Fired
ClickUp just slashed 22% of its workforce to replace roles with AI agents, and tech layoffs in 2026 are already nearly matching all of 2025. Box founder Aaron Levie coined it "AI psychosis" — executives confidently automating jobs they don't actually understand, driven by hype rather than operational reality.
Why it matters: When the people pulling the trigger on headcount reductions have never done the jobs being eliminated, the collateral damage extends well beyond the spreadsheet.
Coders Are Outsourcing Their Brains to AI — and It's Getting Risky
Developers are increasingly refusing to write code without AI assistance, and researchers are warning this is producing faster but not better code. The deeper concern: engineers are losing the foundational skills to catch the mistakes AI confidently ships.
Why it matters: If a generation of coders can't audit AI output, the technical debt piling up in codebases today becomes tomorrow's catastrophic security and reliability problem.
Groq Pivots to Inference, Chases $650M Raise
Fresh off Nvidia's eye-popping $20B "not-acqui-hire" deal activity reshaping the chip landscape, Groq is reportedly raising $650M as it shifts its focus away from hardware and toward AI inference — optimizing how models respond to queries at scale. The round is internal, per Axios.
Why it matters: Groq's pivot is a telling signal that the real money in AI chips isn't building them — it's running them as fast and cheaply as possible at inference time.
VCs Are Handing Series A Checks to 19-Year-Olds, and They Know It's Weird
Three top VCs spoke candidly about the AI funding frenzy, with one joking that a 22-year-old in SF building AI will find a seed term sheet in their inbox — and a 19-year-old might already have a Series A. The piece surfaces real anxiety about groupthink driving valuations, with investors aware they're in a bubble but unable to stop participating.
Why it matters: When the people allocating billions admit the logic has broken down but keep writing checks anyway, the correction — when it comes — will be brutal for founders who raised on vibes alone.
Quick Hits
- →AI-generated fake Black personas are being used to sell Shein dropshipping junk on TikTok — a new low for synthetic influencer grift. The Verge
- →One ex-Meta engineer walked away from AI startup money to just... build a website, and it's actually working out. The Verge
- →A 17-million-device botnet tied to a Russia-based residential proxy network has been dismantled. Ars Technica
- →Microsoft is threatening a security researcher with criminal investigation after they published zero-day exploit code publicly. TechCrunch
AI TLDR