[shiroiushi]

I thought the consensus was that he was extremely skilled, and he's obviously proven himself at making very useful tools for that system. Why not hire such talent and put it to some use? Not all positions need to have a lot of team interaction.

[randomdata]

Why? Why hire him when you can just as easily hire someone else with all the same skills and a more suitable personality?

[shiroiushi]

You don't know someone else has the same skills and can perform. Lots of new hires don't work out (for technical skill reasons) even after passing the interviews; they're always a gamble.

[randomdata]

Well, we know that the person in question has a low likelihood of working out. As before, other companies, like Apple, tried but were unable to make it work. As his personality demonstrates, and as the testing at Google concluded, he is unlikely to be a good fit for such organizations.

Sure, there is a slim chance that Google could have found the right fit for him in the end with the right accommodations[1], but why take the risk[2] when there are others lined up that are far less risky?

Of course there are no guarantees in life, but when playing the odds...

[1] That would have to be invented. Now you are also relying on the implementer at Google, which brings great risk on the employer getting things right even if the worker somehow magically came risk-free.

[2] Especially when the tests designed to try and estimate that risk raised red flags.