Good news. Legislators have begun a two-year study of artificial intelligence to figure out how the government should regulate it. So, they formed a special committee for the task. That’s a nice little tax-burden treat. Wouldn’t it have been cheaper to ask ChatGPT?
On September 16, the Senate Special Committee on Artificial Intelligence gathered in Columbia to start studying AI, the dangers it poses, and what the government should do about it.
Luckily for them, two AI gurus at the University of South Carolina were there and willing to join this meaningful conversation. Professor Bryant Walker Smith teaches law and studies emerging technology, while Professor Homayoon Valafar runs USC’s AI Institute.
Smith went first, and his testimony was very good. If you have time to watch it, please do. He quickly got to the pesky little problem of defining AI in law. AI technology evolves so fast that the definition could go stale by, let’s say, tomorrow.
He explained that the federal government is already playing this game. It created a broad definition of AI and then specifically excluded word processors and other common commercial products. Well, that worked until those same products started adding AI into the mix. Microsoft Word, for example, now incorporates AI technology that can draft large amounts of text from a human prompt.
Smith even pointed out that some AI definitions have already become “anachronistic.”
In plain English, outdated.
This is kind of similar to what ConservaTruth keeps pointing out whenever we explain the dangers in bills. Legislators broadly define something to protect citizens from it. Then that something changes and starts including what they excluded from the definition, sending legislators right back to figuring out how to write another law to fix the problem they created.
It’s legislative tag. You’re it.
Smith also stated, “But for every definition, I’d say why is it useful in the legislation or the other work?” There’s a thought. Figure out why you need an AI definition before writing one into law.
From there, the conversation touched on machine learning, AI in schools and, somehow, self-driving cars. One senator described a Tesla ride where the car tried to follow the GPS into a gully and asked who would be responsible if the driver failed to override it and the car ended up in the ditch. Smith said something to the effect that, you know, humans are supposed to pay attention and intervene. Well, that clears up the mystery of the Tesla in the gully.
Smith mentioned that existing law already has ways to deal with these situations. Negligence, products liability, agency law, and other doctrines can apply depending on who controlled the system, what the company knew, and what went wrong.
A senator said: “I personally believe current law exists to take care of when they misuse AI and harm our citizens.” But again, if you can’t legally define AI, how will you “take care” of abuse, and what harm do we speak of?
The same senator asked whether AI might require “a new canon of law.”
Oh, no. A whole new body of AI law? Will that come with another chapter in our already behemoth Code of Laws? New boards? New committees? Maybe an agency to administer the regulations and another appropriation to pay for everyone administering them?
Careful, we may have found the next taxpayer-funded solution looking for a problem.
Professor Valafar made it clear he likes AI and encouraged the state to embrace it. He warned, “If you get stuck in the weeds, you’re not going to accomplish much,” and advised lawmakers to “Focus policy development on individual cases. Focus on current challenges.”
Interestingly, neither expert presented AI itself as some looming menace requiring our Legislators to rush in with a sweeping new regulatory framework.
The weeds apparently looked inviting because one senator soon asked: “Do you think AI can become sentient?”
We had gone from figuring out how to define AI to discussing whether AI could eventually experience sensations and feelings. The conversation wandered through human consciousness, morality, and spirituality before Valafar eventually brought it back to earth. “For today, my recommendation is to focus on the technology of today.”
A senator eventually explained why all this was necessary:
“I think the crux of the matter and why our constituents in South Carolina care about this issue and are glad we’re doing this is that they want us to get involved before the toaster burns the house down. That’s the crux of the matter.”
Well, there you have it. The people are begging for this.
Wait, which people, how many people, and what exactly are they asking legislators to do? Did constituents call asking legislators to regulate ChatGPT? Claude? Open-source models? Self-driving cars? AI in schools? AI in police departments? Agentic AI? Neural networks? Sentient machines?
Then Smith said, “Trying to anticipate all of the issues now and legislate for them will create certain rules that will be out of date by the time that they are enacted.”
Smith had another warning: “Whether we’re talking about centralized power or distributed power, there are reasons to be concerned about over-empowering companies, over-empowering governments, or over-empowering individual malicious users.”
The free market should enjoy the arrangement too. Large technology companies have billions of dollars to pay for lawyers, lobbyists, and compliance departments. Smaller AI companies don’t. Give both a thick stack of regulatory requirements and watch equality bloom.
Overall, we suspect this hearing was really about figuring out how the government can get control over AI and get taxpayers to pay for it.
Thank goodness they’ve given themselves two years to do it.
Disclaimer: The views expressed in this article are those of the author and do not constitute legal or professional advice. ConservaTruth assumes no liability for any actions taken based on this content. Read more.
Subscribe to ConservaTruth's Email Newsletter for curated insights on South Carolina's legislative activities and conservative viewpoints, delivered straight to your inbox! With vetted and easy-to-understand information, our newsletter empowers you to become an informed and engaged citizen, actively participating in safeguarding our cherished Constitutional values. Don’t miss out on crucial updates—join our community of informed conservatives today!
Comments