Donate
  • Freedom
  • Innovation
  • Growth

Don't Regulate AI -- at All, for Now

MailChimp
Artificial Intelligence (AI) is the latest technology “thing” that is going to change the entire world, unless it destroys it.

Or at least that’s the impression most people have after hearing about AI in major media and on social media channels.

The truth is, AI is many different things, and it’s not all that new. In fact, in early 2019, the (first) Trump administration issued an executive order on artificial intelligence—three and a half years before the debut of ChatGPT alerted us all to the astonishing abilities of a large language model.

Like many, I remember how impressed I was the first time I asked ChatGPT to write something for me. Fast-forward to today, and I use it regularly to check grammar and spelling when I write, to correct usage and style, etc.

Every time something new comes along, there are immediate calls for government to regulate it. Innovation seems to unnerve people, who are concerned that they are going to lose their jobs, or in the case of nanotech or AI, that it somehow is going to extinguish the human race. And politicians capitalize on these fears by promising to protect their voters from the scary thing.

Virginia Postrel has been describing the tension between dynamism and stasis in her book and on her blog since 1999. There is a perceived safety in stasis—the way things are now—and a perceived risk in dynamism. But dynamism, change, innovation, revolution, is how society advances. At various times society has been afraid of the Industrial Revolution, of electricity, of wireless technology, of vaccines, of nuclear power, of robotics, of the Internet, of biotech, of nanotech, and now of AI.

But all of these things have improved the human condition, extended the length of life and improved the quality of life, reduced poverty and drudgery, and created opportunities. Nonetheless, the fear of dynamism continues.

Of course there are risks to new innovations. And when it has become clear that a technology has caused harms along with its benefits, regulation is absolutely appropriate. The danger is regulating too early, which risks precluding the benefits of innovation.

This second tension, the tension between precaution and permissionless innovation, has been described by our friend Adam Thierer. Our bias should be toward permissionless innovation, not precaution. Precaution is a recipe for stasis—for a society that has abandoned progress.

Right now, governments at the state level are moving bills to regulate AI that should be roundly rejected. Not only do they violate permissionless innovation, but it’s ridiculous to think that a technology should be regulated on a state-by-state basis. AI isn’t exactly intrastate commerce.

IPI Resources on AI:
  • A Rush to Legislate on AI Is a Bad Idea
  • Podcast with Dr. Dan Garretson on Gemini AI
  • PolicyBasics podcast with Dr. Dan Garretson on the basics of AI
    __________________________________________
Today's TechByte was written by IPI President, Tom Giovanetti.