Three stories crossed my feed today and they line up into one arc: the money behind AI, the machines it's moving into, and what it's doing to jobs. Here's the builder's-eye view.
1. OpenAI spent $34B and made just $13B
The Financial Times verified OpenAI's audited 2025 financials ahead of its IPO: about $34B spent (roughly $19B on R&D, nearly $6B on sales and marketing) against about $13B in revenue. The headline net loss was ~$39B — but ~$30B of that was a one-off, non-cash charge, so the real operating loss was closer to $8B.
Why it matters if you ship on these APIs: the low per-token prices we all build on are being subsidized by investor money. That's great for your margins today, but it's not a law of nature. If your unit economics only work at current API pricing, model out what happens if it climbs — and keep an open model in your back pocket.
Source: Reuters/FT via Investing.com · Ed Zitron's analysis
2. Humanoid robots hit mass production
Figure's humanoids finished an 11-month pilot at BMW's Spartanburg plant — over 90,000 parts handled, contributing to 30,000+ vehicles built. Figure 03 is now ramping production and being evaluated for BMW's European plants, while Boston Dynamics started shipping its electric Atlas to Hyundai.
The interesting shift for builders: physical AI is moving from "cool demo" to repeatable production deployment. The software stack around these robots (perception, planning, fleet orchestration) is the same agent-and-tooling problem many of us already work on — just with a body attached. Worth watching where the SDKs land.
Source: Figure — production at BMW · BMW Group
3. AI's job impact is uneven — cuts AND hiring
2026 tech layoffs passed 142,000. But OpenAI's Sam Altman says the "jobs apocalypse" he expected hasn't arrived ("delighted to be wrong"), and firms that actually adopt AI tools are hiring more. Gartner found ~80% of execs cut headcount to fund AI — yet augmenting staff tends to beat replacing them.
The takeaway I keep coming back to: the split isn't "AI vs. humans," it's people who use AI vs. people who don't. As a builder, the leverage is in being the one who ships faster with these tools, not the one waiting to be automated.
I run a small daily AI-news channel and write these up as I go — still figuring out the workflow, so feedback's welcome. If short daily recaps are useful, the video version is embedded above.