The NYT article (link to which is broken) provides interesting context:
That was the initial offer that was turned down. Reading between the lines, maybe another reason for the stratospheric salaries is that nobody actually wants to work for those companies.
I'd also point out an often missed aspect:
There's nowhere else in the world, especially academia, that even comes close to being able to give you those kinds of resource promises. Even if you know their AI hype is BS, if you want to do cutting-edge research you need to play ball if you want access to resources.
These companies know that they have only two assets: lots of compute resources and eye-watering compensation packages. Nobody would be working for them otherwise.
The whole "superhuman AGI" stuff is also not really a required motivation. This is an existential crisis for these companies just as much as it is for the creatives and other people they're putting out of work. If Google's search product is nothing but an AI chat bot, then once a slightly better AI chat bot appears Google is out of the game.
This is an extremely important point. There's as much a chance that they're trying to capture research that would be going on out in the open and keep it to themselves as they're trying to research AGIs. Consider what Google or Meta is worth when that AI chatbot can just run locally on anyone's machine without ads. These are fear based moves, but not for the aspirational reasons they're stating.
.png)

![What the hell is wrong with Europe? [video]](https://www.youtube.com/img/desktop/supported_browsers/opera.png)
