Inside the Fight Over AI Regulation, China, and America's Future

**SPEAKER_1** (0:00)
This is a monday.com ad. thesamemonday.com helping people worldwide getting work done faster and better. thesamemonday.com designed for every team and every industry. thesamemonday.com with built-in AI, scaling your work from day one. thesamemonday.com that your team will actually love using.
thesamemonday.com with an easy and intuitive setup. Go to monday.com and try it for free.
Yes, the same monday.com.

**Michael Smerconish** (0:30)
Welcome to The Smerconish Podcast for Independent Minds.

**Michael Smerconish** (0:35)
So in the last hour of the program, I was chatting with Jonas Serra, the subject data centers. That piece that he wrote is in the smerconish.com daily newsletter today. There's something related in the newsletter today, and that's an essay from Mark Zuckerberg from the Wall Street Journal, which begins this way. We're fortunate to live at an incredible moment in history. In the next few years, people will be able to use superintelligence beyond human capacity to create and discover extraordinary new things, build new businesses, express our ideas, learn new concepts, and improve our lives, health, relationships, and careers. He said, the defining question of our age isn't whether superintelligence will exist, but who will have access to it. And then Zuckerberg goes on to essentially argue that it ought to be distributed broadly to individuals, not concentrated in the hands of governments or a few corporations, because Mark Zuckerberg argues widespread access will maximize innovation, it will preserve liberty, create prosperity, etc., etc., etc.
I guess the counterargument, if I were looking for one to make to Zuckerberg, would be that AI is not analogous to any prior technology or development, because it risks scale much more quickly than the institutions that are designed to contain it.
If you're having difficulty following this conversation, and I certainly am, I'm struggling to keep up and to understand, but I'm keenly interested, then, as Ron Popiel would have said, but wait, there's more, because also, in the newsletter today is a great essay written by Richard Clarke, who, of course, was the nation's first cyber czar.
It's titled Demythosing, Demythosing, as in mythos, you get the reference? The AI. Debates on Regulation. And Richard Clarke is here to help us understand and keep pace with this conversation. Thank you so much. I know I'm pulling you out of a commitment. This is really important. And I thought it was so great and funny the way that you wrote it. You say regulating AI was not the Trump administration's agenda initially. Their tone has started to change. Set the stage before I ask you about the four terms that you define.

**Richard Clarke** (3:04)
Well, the Biden administration was setting up an AI safety and security process in the Commerce Department with experts. The Trump administration came in and blew all that up and said, we don't want to kill the baby in the crib, the baby being the enormous AI industry, some baby. We don't want to pick winners and losers. We don't want to regulate. We just want to do nothing that will stop AI. We have to win the race with China. So that was their position for the first year. And then one day, a few weeks back, on a Friday afternoon, they said, wait, wait, we're going to regulate Anthropic's new product, Mythos, and we're giving Anthropic 90 minutes to set up a system whereby only Americans can get access to it.
So they stumbled into regulation and they stumbled when they did it.
But there are a number of issues which they're now debating inside the Executive Office Building, inside the West Wing, inside the Commerce Department, about whether or not to regulate additionally or in some way. Issues like, should we stop the Chinese AI programs from being sold in the US? Should we stop open source, open weight systems? What should we do about guardrails in these programs? And we don't know where they're going to come out. They are actually debating. They're actually factions in the administration. But the problem is they want to do some minimal regulation and they don't have a regulatory framework in mind.

**Michael Smerconish** (4:50)
Do we even have the ability, we as the United States, to set this stage to determine the parameters or the guardrails? What control do we have over the axis of evil or anyone else on the planet who seeks to develop artificial intelligence?

**Richard Clarke** (5:07)
Well, we are usually the international lead. The British are actually taking a lead in AI safety right now because we're not.
We certainly have some control over what goes on in the United States and what's sold in the United States, what's used in the United States. We might be able to say that nothing should be sold without guardrails that work. What's a guardrail?