Do your colleagues at work have access to all of your own accounts, or are they expensive tamagotchis?

I trust that my colleagues are not going to do stupid things with their accounts, I cannot say the same for agents, and making their own accounts that you’re still responsible for seems like you’ve just moved the problem rather than actually resolving it, since ultimately, you’re still responsible for their actions.

Are agentic systems like these making correct decisions more often than humans really unthinkable? A comment like what you just wrote was unimaginable 5 years ago

Humans can take accountability for mistakes, and there are systems in place to help you if they don’t.

I imagine the same systems will evolve for agents as well, if nothing else because lack of trust will impact provider's bottom line.

How so? Each human is a different "model", and its constrained to the physical world. What are we going to do? Put whole corporations in jail, shareholders included? Limit what they can do?

The "it went sideways" scenario for 100k agents spawned across the world using the same bad model is completely different from humans going sideways.

The power imbalance will ensure that this is unlikely to happen, in much the same way that people generally don’t trust that massive corporate entities have their best interest at heart but might have some regard for their colleagues wellbeing.

Work environments usually already have strict controls on financial accounts for very obvious reasons. This does result in stupid "why do I need director approval to buy a stapler?" stories, but the alternative risks all sorts of internal and external frauds.

(a notable achievement of SaaS and now AI has been to totally circumvent spending controls. You might not be able to spend $10 on a USB cable without a purchase order, but you can run up an AI bill of arbitrary size and in some places are encouraged to!)

when my colleagues do something illegal or negligent they are personally on the hook for it. Who is on the hook when my bot does that?

You just give them the permissions they need to do their work.

This approach exploits the fact that managing permissions for humans is a very old requirement and most platforms have good support for it already. You can then issue API keys from the agent's accounts if you want to restrict permissions further.