The real existential risk isn't AI. It's us.
Everyone's warning that AI is "existential." Here's the opposite case — built entirely on numbers you can check.
Read the essay →"We hold a four-day lab experiment to a higher standard than we hold ourselves."