You might like Anthropic. You might think they're the good one. They might even be reaching the conclusion that they don't want, say, an authoritarian government shutting them down capriciously in good faith. But this is actually pretty illustrative of why a for-profit race to superintelligence just doesn't work: push comes to shove their incentive is to block federal limits on their work. It's why we need government to set the guardrails here, not the AI companies. xitter.cf/politico/status/211121…