Episode 51: AI Layoffs Go Explicit, Codex Goes Agentic, and Chips Eat $20B
· 11:19
Show Notes
- Snap lays off 1,000 (16%) and says AI generates 65%+ of new code, targeting $500M annual savings
- OpenAI ships major Codex update — 3M+ weekly developers, deeper "computer-use" agentic workflows
- Anthropic releases Claude Opus 4.7 — same token pricing, prompt caching (up to 90%) and batch processing (50%)
- OpenAI reportedly commits $20B+ over three years to Cerebras chips; Cerebras files for IPO
- China AI capital markets: DeepSeek raising at $10B valuation, StepFun preparing IPO
- AI Tool Spotlight: Mokka — AI sourcing and screening across 850M+ profiles and 250+ job boards
- Hiring insight: add a 30-minute "AI leverage" station to your onsite loop
Key Takeaways
- "AI productivity" is now a board-level lever for headcount — expect fewer, more senior builders who ship with copilots
- Agentic developer tools are becoming default — hire for AI leverage, evaluation discipline, and security awareness
- Inference-first infrastructure is being funded aggressively — performance, distributed systems, and hardware-aware ML roles spike
- Source aggressively from layoff pools (Snap, Disney) — top candidates disappear fast
- Run a third-party access audit for recruiting and developer tools — vendor tooling is a real attack surface
Funding Watch
- Artemis — $70M, cybersecurity (anti AI-powered attacks)
- Nava — $8.3M seed, guardrails for AI financial agents
- Lua — $5.8M seed, "Agent OS" for human-agent collaboration
- DeepSeek — raising at reported $10B valuation
Frequently Asked Questions
What does Snap's 16% cut and "65% AI-generated code" claim mean for hiring?
It signals AI productivity is now a board-level justification for headcount cuts. Expect hiring to bias toward fewer, more senior engineers who can ship safely with copilots and agents, plus platform and reliability roles that keep AI-assisted delivery stable.
How should I interview for "AI leverage"?
Add a 30-minute live station where the candidate uses their preferred AI tool on a small real task (bug triage, test plan, PR review, refactor). Score reasoning, validation discipline, security awareness, and output quality — not raw keystrokes.
Why is the OpenAI–Cerebras deal a hiring signal?
A $20B+ multi-year commitment plus a Cerebras IPO filing means inference-first infrastructure is being funded at scale. Expect hiring spikes in inference performance, cluster and systems engineering, and large-scale networking and observability.