Subscribe
Share
Share
Embed
What's your AI success story? What made you realize you had to change? And what does a hopeful future with AI look like to you? AI Hope with Geoff Livingston asks three simple questions of people doing the hard work of adapting and thriving in the AI era.
Geoff Regulation
===
[00:00:00] Hey everybody. My name is Jeff Livingston and this is AI Hope. My book, Now Is Gone, is about never-ending change. This podcast is about what comes after the stories, the people, and the reason for hope.
Hey, welcome to the most recent edition of AI Hope. And today, I wanna talk about regulation. And the reason why is obviously a lot of people want regulation to happen of the AI industry, and that's because there's been a whole host of things that have happened that basically cause us not to trust AI companies.
And so with that in mind, ~you know, ~whether it's cybersecurity challenges, data centers, flock, ~um,~ job displacement, dismay with frontier models, AI slop, ~uh,~ IPO hype, ~uh,~ whatever it might be, there's a lot of AI needs to be reined in. So how likely is that to happen? First of all, given the United States and our current [00:01:00] regulatory environment, and also where will it happen and when can we expect that to occur?
I did write a substack on this. I'll include it on the show notes if you wanna go read in depth. But in general, I would say on a federal level in the United States, I would not expect any regulation of substance to occur, ~uh,~ within the next two years. And that's because of the current political environment.
We're all aware of the situation, regardless of which side of the fence we sit on. ~I mean, ~clearly Congress is not functional. ~Uh, ~people aren't working together to find common solutions. There's a lot of disinformation out there as well, and it's all creating massive issues for, ~uh,~ folks that are trying to get things done.
~Uh, and, ~and so whether it's the administration itself trying to force regulation by fiat, ~uh,~ or through its various agencies, ~um,~ [00:02:00] or it's Congress, I don't think you're gonna see anything beyond just ~kind of ~like these haphazard things where we've seen, like Mythos, for example, by Anthropic, their frontier model delayed for a few weeks, or the administration, ~um,~ telling people what they can, ~uh,~ produce and not produce as far as, ~uh,~ algorithms that'll work with the Department of Defense, and you're seeing that also with Anthropic.
~Um, ~there are a few reasons beyond just the dysfunction. ~Uh, ~one of those reasons is actually the big tech players are pretty prevalent in lobbying Congress right now, and you can see that with the way some of these regulations to date have been written and the amount of influence that we see from these tech moguls that are interacting with Trump.
~Uh, ~and then the other thing that can also really stop regulation from happening, particularly f- directly from the administration itself, which isn't necessarily incentivized to do it, as we know this administration try to phrase my words carefully, lean towards [00:03:00] supporting, ~um,~ corporations as opposed to individual freedoms.
~Uh, ~and we've even seen that with privacy and IP online, where they basically feel like LLMs have the right to scrape anybody's content to train. ~Um, ~but the entanglements that occur afterwards in the justice system and all the different court cases and, ~you know, ~the orders are getting struck down or delayed or escalated up to the Supreme Court, it's a hot mess.
That being said, Europe is doing quite a bit of regulation right now. ~Um, ~they are looking at AI. They are also looking at some of these companies, and we've seen that already with Anthropic, ~uh,~ and the whole, ~uh,~ disclosure rule. So in short, the EU basically mandated that any AI content creator watermark or disclose that its content was created with AI, and Anthropic immediately complied by putting, ~um, uh, ~what I would call a, ~uh,~ quick, ~uh,~ watermark into any of its [00:04:00] content, even if it's just edited by Anthropic.
So there's been a lot of disruption about that. ~Uh, ~Substacks are providing some sort of, ~uh,~ AI awareness, so there's a lot of different things in there. Please excuse me. Take a little quick drink here
So in many ways, just like with privacy, the EU is leading the way. But then there's also state and local regulation. We're seeing, ~uh,~ I think more than fifteen hundred bills to regulate AI have been introduced. Now, obviously, bills can be just political statements, and some of them do get passed, though, and we are seeing serious legislative efforts in California.
I think Texas is another state where they're doing stuff. ~Um, ~and then on the local level, you see all the data center fights. And then finally, the last area of regulation is when individuals stand up for their rights and then take these frontier models and these courts, ~uh,~ to, ~uh,~ court and basically sue them and-- to protect their rights.
And these civil cases have been ~kind of ~effective. ~I mean, ~they've reined in Meta [00:05:00] with all of its algorithmic work in Instagram and Facebook that's harming teenagers. Anthropic was called to task for basically stealing content from authors like me. I was in one of those cases for Marketing the Round, which is above me right here.
~Um, ~so when you see things like that, ~uh,~ you definitely have faith that maybe some parts of the government are working to protect people. So in a nutshell, is there hope for regulation? Yes, but probably not on the federal level, at least over the next couple of years. ~Uh, ~the good news is that a lot of voters are ~very, ~very angry, ~uh,~ that they're letting their voice be known, and that politicians on both sides of the aisle are recognizing the need to, ~uh,~ regulate.
So hopefully we'll see more of that. And maybe, ~I mean, ~who knows? Maybe there will be enough political will to actually engage something on the federal level. Wouldn't that be something? ~Um, ~so also, I would say if you're corporations, another thing to consider is how are you gonna [00:06:00] implement AI, knowing that regulation's probably going to arrive either through Europe, ~uh,~ a big state like California, ~uh,~ judicial fiat, governmental fiat.
Somehow, some way, some best practices are gonna touch you. ~I mean, ~obviously, the more mindful you can be about your employees, your customers' personal information, avoiding surveillance type of apps, ~or,~ or at least getting permission to access data and use that data, that's gonna be the home run, I believe.
So with that, thanks for listening to this edition of AI Hope. This is a short one, but I think it's still a, a useful one. Cheers.