00:00:00
I think it's very concerning. There's a regulatory frenzy happening at the states right now. Let me give you some
00:00:05
stats on this. All 50 states have introduced AI bills in 2025. There's been over a thousand bills in state
00:00:12
legislatures. 118 AI laws have already been passed across the 50 states. Everyone just seems to be motivated by
00:00:19
the imperative to do something on AI, even though no one's really sure what that something should be. Exactly.
00:00:25
And there's no real agreement on like what all these AI regulations are supposed to do. So they're just making
00:00:29
things up or what the risks are or what the risks are. So you've got 50 different states each with their own
00:00:34
reporting regime, which is going to be a trap for startups. They've all got to figure this out about what they're
00:00:39
supposed to report on, what the deadlines are, who to report to. And if you want to see where this is going,
00:00:44
look at Colorado. This has already been passed into law, SB24-205, Consumer Protections for Artificial
00:00:51
Intelligence, that bans something they call algorithmic discrimination. Algorithmic discrimination is defined as
00:00:57
unlawful differential treatment or disperate impact based on protected characteristics. So things like age,
00:01:04
race, sex, disability. If any of those factors drive an AI decision and it results in a disparate impact, then both
00:01:12
the developer of the AI model and the deployer, which means the business that's using it, can be in violation of
00:01:18
this law and they can be prosecuted by the Colorado attorney general. So, let's say that you got someone like a mortgage
00:01:24
loan officer who's reviewing applications and let's say they don't even discuss race. They're just using
00:01:30
race neutral criteria like a credit rating or financial holdings. If the result of their decision nevertheless
00:01:36
had a disperate impact on a particular protected group, its decisions could be found to be discriminatory and moreover
00:01:44
the developer of that model could be liable even though their model just gave an answer that under the circumstances
00:01:50
was truthful. The only way that I see for model developers to comply with this law is to build in a new DEI layer into
00:01:57
the models to basically somehow prevent models from giving outputs that might have a disperate impact on protected
00:02:04
groups. So, we're back to woke AI again. And I think that's the whole point.