The future can't belong by default to whoever builds it first.

You can't leave. At Code for America we're testing this with Anthropic: a SNAP tool for caseworkers only counts as a win if fewer eligible families lose help to paperwork, if people can still challenge wrong decisions, and if caseworkers keep real authority to overrule the tool.

These warnings shouldn't push government further from AI. They should push it to be better prepared to understand it, challenge it, and refuse it when needed. Government carries a weight most companies don't: if an app fails you, delete it. If you need SNAP, there's no other agency to switch to.

My phone's been full of AI fear: a former OpenAI/Anthropic researcher warning companies are "gambling with our lives," an Anthropic alignment lead putting extinction risk above 10% in a decade. I run Code for America, and I build with Anthropic. Here's my take: 🧵

"Work requirements" sound simple until you meet the people they hit: workers cycling between jobs, not sitting on the sidelines. New CalFresh rules cap benefits at 3 months every 3 years for them. The real fix isn't more red tape, it's making the system actually usable.

We need everyone to speak out. We can’t stay silent while so many of us—Latinos, Hispanics—are carrying our passports just to feel safe. Donald Trump's immigration policy and lack of due process doesn’t just target immigrants; it puts every brown person at risk.

This is exactly what we've been saying for months. DOGE is making it harder, not easier, to run the federal government. Our seniors deserve so much better.