>>106289751
Very true, but with AI specifically they have to censor it all the time because it's the antithesis to how they operate. By design they'll never be able to use AI without gimping it.
Plus such things will backfire on them in ways they can't imagine. They'll have robotic bodyguards that get hacked and off their owners with the security tapes wiped so the culprit can't be identified. They'll perfect a shopping experience only to have no customers for that brand.
They'll perfect farming only to have people support the mom and pop stores.
It'll be a competition and they'll shoot themselves in the foot every step of the way.
It only takes one mistake too.
>machine "hallucinates" once or twice
>Am I going to trust this machine to do x task for me properly?
>HOTFIX BUG FIX
>...I need to be able to trust something out of the box. Sorry.