Technology
Meta Cuts 8,000 Jobs as AI Spending Surges
Meta will cut 8,000 jobs May 20 while pouring tens of billions more into AI. The 'AI replaced them' headline is neater than what the company actually said.
AI
Healthcare IT
ChatGPT for Clinicians Is OpenAI's Point-of-Care Play
OpenAI's new clinician-only workspace is less about replacing doctors than owning evidence review, documentation, and prior auth at the point of care. The opportunity is real. So are the governance boundaries.
AI
ChatGPT in CarPlay, Ten Months After I Asked
I emailed OpenAI a feature request in June 2025. This morning, on my drive to work, I used it.
AI
NSA Is Using Anthropic's Mythos Despite the Pentagon Blacklist
Axios reports the NSA is using Anthropic's unreleased Mythos model — the one Anthropic says can find zero-days autonomously — despite the Pentagon formally labeling Anthropic a supply-chain risk. Here's what Mythos is, who else has access, and why the operational need is winning.
AI
Stanford's 2026 AI Index: Twelve Months, One Jagged Line
Stanford's 2026 AI Index has the receipts for what this year felt like: agents jumped from 12% to 66% on real computer tasks, the US–China gap narrowed to 2.7%, and the best coding models still can't read an analog clock.
AI
OpenAI's Codex Moves to Your Desktop
On April 16, OpenAI shipped a Codex update that lets the agent see your Mac screen, move a cursor, and type across any app. This is the answer to Claude Cowork — and it reshapes the desktop-agent race.
AI
Claude Code Is a Lot More Than a Chatbot
Most people still picture a browser and a chat window when they hear 'I used Claude.' That mental model is three years out of date.
Healthcare IT
Enterprise AI Plans Are Harder to Get Than You'd Think
Two months into pricing a small Enterprise AI plan with a HIPAA BAA, one vendor has answered cleanly, one has answered with meetings, and one hasn't answered at all.
AI
Claude Opus 4.7 Arrives. Mythos Still Waits in the Wings.
Anthropic released Claude Opus 4.7 today — better coding, sharper vision, a model that checks its own work. Its best model still isn't shipping.
AI
Tokenmaxxing Is Burn Rate Theater
Measuring engineers by how many AI tokens they burn is the lines-of-code metric wearing new clothes — and with compute now a genuinely constrained resource, the waste isn't free.