Joe Allen

Yes, Steve.

If the posse takes anything away from today's show, duty of care.

This is a legal concept that if a company is putting a product out, then they have to run risk assessments to see if there are any

predictable harms that their product could cause.

So for instance, if you have a chat bot like ChatGPT and you know that it might seduce children sexually or seduce children into committing suicide, then you are then liable

for eliminating that risk before deploying the product?

Well, we know right now that ChatGPT does in fact do this.

They know that the product does this.

And so in the absence of any legal framework or any kind of law or regulation that would hold them accountable, they get away with it.

So Marsha Blackburn's Trump America AI Act, in the very beginning of it, third section,

They say that duty of care will be first and foremost.

Any company who is producing any kind of algorithmic system that they know could get kids addicted, could lead to AI psychosis, could lead to suicide, they will be liable for the harms done.

and they will be responsible for eliminating the harms before deployment.

I think this is one of many positive steps that government could actually provide in order to throw sand in the gears of the machine.

Well, I would send them first to joebot.xyz.

I've got an analysis, a review of Marsha Blackburn's Trump America AI Act.

I tried to break it down, put it in layman's terms and bring out the most important elements of it and also some of the problems that I foresee.

Also, my social media at joebotxyz.

And Dark Aeon, unfortunately, the signed copies are on hold while I'm traveling.

You can find it anywhere books are sold.