If you're choosing amongst those companies, you're in the very fortunate situation of being spoiled for choice. It's a wonderful position you've found yourself in.

Some related thoughts:

• Choice of company, when each one is so stellar, is often not as important as choice of team or role. Or, for that matter, even matching yourself with a great manager.

• The fates of OpenAI and Anthropic are going to be higher variance than Meta (or Microsoft or Google). This doesn't mean they're bad choices -- it just means that predicting their futures comes with bigger error bars.

• The fact that you describe OpenAI and Anthropic as "more exciting" is probably a hint you should at least strongly consider joining one of them simply because your interests point in that direction. There's a lot to be said about being excited about something.

• When a choice is too close to call, don't overthink it. You could be aiming for false precision. In cases where it's simply impossible to know which path is best, you're going to have to choose based on available information, resting assured you could not have made a better choice at the time given what was knowable.

Wishing you the best in your decision.