A.I.’s sneak attack
A provision buried in Trump’s slash-and-burn budget could liberate the bots and doom the humans. And more on the topic
Buried in the One Big Beautiful Bill Act — the budget that the House of Representatives passed last week — is a provision that would forbid the states from passing any regulation of A.I. for the next decade. Given the White House’s close relationship with A.I. movers and shakers like Elon Musk and OpenAI head Sam Altman, and the failure to pass any comprehensive federal regulation of A.I. thus far, the industry may be set to enjoy unfettered growth in a regulatory wilderness for the foreseeable future.
And that should give everyone pause. This is happening just as generative A.I. appears set to upend the economy as we know it, with 40 percent of employers polled worldwide by the World Economic Forum this year saying they planned to reduce their workforce to take advantage of A.I. automation.
This afternoon at 12:30 p.m. Eastern, we’re going to be speaking Live with the journalist and author Karen Hao about her new book, Empire of AI: Dreams and Nightmares in Sam Altman’s OpenAI, which explores in detail how the pursuit of artificial intelligence took over Silicon Valley, how it's already changing work and lives around the world, the implications for the future, and what can be done about it.
Before today’s live event, we invite you to look back at our past conversations about A.I. and its impacts with Hao, the activist Sneha Revanur, and the author R.O. Kwon, and the short discussion that follows with the technologist Catherine Bracy, taken from our conversation earlier this year about the ideology of venture capitalism that underlies the business models and world-changing dreams of the oligarchs, the direction of tech innovation, and, increasingly, the future we all face.
So what’s the role of A.I. in the future of the economy?
The impact of A.I. on the labor market is a big deal. It is potentially catastrophic. It's the next step beyond the marketplaces, and beyond how the idea of the marketplace has made everyone an independent gig contractor.
Is the worry that A.I. will put everyone out of work entirely?
It's like 100 times the risk that the gig economy is in my mind. One of the concerns that I have is that people in my community of responsible tech policy, people on the left, aren't taking A.I. seriously. They think it's vaporware.
And I think that's a dumb argument, because it doesn't matter. It's the same negative labor implication. It doesn't need to be general intelligence. It doesn't have to be A.G.I. for A.I. to be harming workers. It's harming workers today in its current form.
People discount the risk. They say this isn't going to put people out of work. It's going to augment their job, or at worst it's going to surveil them or going to make their job worse or it's going to drive wages down.
I think it's going to displace a lot of workers. I think a big thing that's holding companies back from adopting A.I. in a lot of roles is just the friction of changing the status quo.
Though this is where it's very coherent with the Trumpian scorched-earth strategy, the willingness to just throw everything out.
I think the Elon Musks of the world really believe that A.I. could do a lot of stuff. I don't think it would work very well. But anyway, they believe that.
If there is a recession, which is looking like a strong likelihood now, companies start shedding jobs, if we thought the recovery from 2008 was bad for workers… I think this recovery is going to be terrible. Because it's all going to be agents and robots who come back to the jobs. Companies are going to accelerate the adoption of A.I., even if it doesn't work well.
The thing that I keep thinking about is CVS. We're going to lock all the stuff behind Plexiglas, and you have to ring a button if you want a worker to come open it for you. They won't hire more workers. Instead of hiring more workers, they've created an app. And you have to download the CVS app and go in and do facial recognition, or give away all your personal data in order to open the thing. These companies do not care about creating a good customer experience.
So the fact that the technology is not good yet, that it's not as good as a human yet, is not a reason that they will not replace a human with it.
Exactly. And another thing with the drugstore apps is that the retailers look at this kind of customer pain point as another marketing opportunity. It's another point of touch where they can collect more data.
Yes, I was going to say they can get more data from you. We’re working on price-fixing regulations right now. It is currently possible for companies to dynamically price — they call it dynamic pricing — based on time of day, etcetera. Say you're a tired parent who ran out of diapers, and you have to go to the store at 1:00 a.m. to buy diapers. Facial recognition. I see you, I notice that. I see what time it is. I have all your shopping data already.
I'm charging you more for these diapers because I can see that you're a tired parent who is probably willing to pay a premium for these right now. That is possible today. So in all of these ways, I think a recession would accelerate the adoption of A.I., even if it doesn't work well, and that is really scary to me.
And in that sense, I am also a doomer. But I'm not a doomer because I think this technology is going to work really well and be like sentient and take over the world. I know how corporations operate, and that's what I'm very pessimistic about.
If you think about Sam Altman and people like that, they're banking on a world transformational product, but A.I. is not necessarily a machine that's going to solve climate change. It's likely to be a machine that makes it possible to more easily charge tired parents more for diapers the later they go to the store.
If you look at what they're selling, the business model for all the A.I. companies is about automating work. That's what they have found. It's not ChatGPT subscriptions. That's not where the money is to be made.
Can you — and how do you — put that genie back in the bottle? Because it has a lot of other negatives. Also, it's going to use up all the fossil fuels, and then it’s going to completely set back any kind of attempt to mitigate climate problems.
Yeah. There’s a column of worry around this question, and the question of what economy we want to build after this destruction is over.
The A.I. economy seems bad for everybody: bad for workers, bad for the planet, and does stuff that nobody wants. It's like Wario capitalism.
And that's the thing — it isn't about the technology. A.I. on its own could be super powerful. I am somebody who has a genetic mutation that means I will have a 30 percent chance of having Alzheimer's disease by the time I'm 80. I want A.I. deployed on finding solutions faster. Every day, I'm like, I hope I've got 20 years left for them to discover a cure for Alzheimer's. I want that to happen.
I want the conditions to exist for science and technology to develop in a way that is going to address those real problems. But the capital structure around these companies right now is not optimizing for that. It is optimizing for, "Oh, I see that the business opportunity, the way to make the most money, the way to pay back my investors, is to automate work, is to dynamically price diapers."
That is where I think the real concern is. How do we shift those incentives around how investors are investing in A.I.? We need to figure out how to change that investor dynamic if we're going to address the A.I. issue. It might end up being too late, but it's definitely the thing that is shifting the technology on a really bad course
I talked to Sam Altman for the book before he got fired. And he said very clearly, then, that OpenAI should not be a for-profit company. This should not be something that investors are driving the conversation on.
And look where we are now. That was very terrifying to me. I know. He knows.
Live conversations this week!
Join us today, Tuesday, May 27, at 12:30 p.m. Eastern, when we’ll talk with author and journalist Karen Hao about her new book, Empire of AI: Dreams and Nightmares in Sam Altman’s OpenAI. Then on Wednesday, May 28, at 12:30 p.m. Eastern, we will meet Live with The Ink Book Club, and on Thursday, May 29, at 12:30 p.m. Eastern, we’ll be back with messaging guru Anat Shenker-Osorio.
To join and watch, download the Substack app (click on the button below) and turn on notifications — you’ll get an alert that we’re live, and you can watch from your iOS or Android mobile device. You can also watch by visiting our homepage. And if you haven’t already, subscribe to The Ink to access full videos of past conversations and to join the chat during our live events.
It seems there may be MANY things buried in this monstrosity that too few people are talking about and even less are aware of and even less than that give a shit about.
In my younger days when life was simpler and the trees were greener, we never worried that our elected government officials would not ultimately do the right thing. We could live our lives mostly enjoying the blessings of liberty and only needed to think about politics every so often when an election was rounding the corner. Sure, there were always those that grifted along the way and dutifully resigned in shame when they got caught with their hands in the cookie jar.
Now, the grifting is the norm and done in broad daylight. We have to keep our eyes peeled. We bounce all around the internet finding new ways that the comfortable rug our liberties is being yanked out from under us. Our second last line of defense, the courts, have now admitted that they don't trust the U.S. Marshalls, the only people that can enforce their rulings.
Now I only have, between myself and my liberties, my LAST line of defense.
And it is me.
"To Ourselves and Our Posterity"
I watched a brief discussion of Karen Hao's new book, *Empire of AI* on MSNBC. She knows the subject well. I suggest we connect with the Zoom today. Major trouble if not regulated by government.