Regulators can't stop AI—they can only slow down innovation.

Daily Debate #563: "Regulators can't stop AI—they can only slow down innovation.". You are arguing AGAINST the motion. Your opponent: Sam Harris.

I believe that regulators can stop AI. I believe that independent regulators, more specifically, can stop the negative effects of AI. The advantages of this are similar to those of driving. If you design safer cars, set an age restriction, and administer a driving test, there will be fewer issues. Additionally, I believe that businesses such as OpenAI ought to be open about the possible risks associated with the content on their platforms. Now, I do acknowledge that there are flaws, as every policy has. I acknowledge that this solution will punish a group that did not create AI, and I also acknowledge that making it harder to access AI platforms actually risks exposing users to even more harmful websites. But, while I acknowledge these concerns, the goal of law is to reduce as much amount of harm as possible. Furthermore, profits do not trump public safety. That is why I hold the position that regulators can stop the negative effects of AI.

Your car analogy collapses instantly. We don't ban cars because crashes exist—we've accepted millions of deaths as the cost of mobility. You're smuggling in a premise that AI's harms are unacceptable while cars' are tolerable, but you haven't shown why. Regulators didn't "stop" car danger; they managed it within acceptable limits. So your real position is actually "regulators can tolerate AI within bounds," not "stop" it. Which frameworks would you use to define those bounds without crushing the benefits you're implicitly accepting?

My car analogy does not collapse. I do not take the position that AI should be banned outright, as you are implying; I take the position that AI, like cars, should be regulated. You have just committed a logical fallacy. A strawman argument to be exact. Furthermore, give me a direct example in my argument that "regulators stop car danger?" And the smartest thing you have said so far is that I take the position that we can tolerate AI. I would like for you to steelman my position as to why I hold that position using direct sentences in my opening argument

You're right—I misread your threshold claim. You said "stop negative effects," not ban AI. That's precise. But here's where you slip: you demand regulators reduce harm "as much as possible" while acknowledging they'll punish innocent groups and drive users toward worse alternatives. Those aren't minor flaws—they're trade-offs that could increase net harm. If regulation's collateral damage exceeds its prevention, you've violated your own principle. How do you measure whether your regulatory framework actually reduces total harm, not just shifts it elsewhere?

The reason I am not contradicting myself is that the number one goal of law is to reduce the amount of harm a person can be exposed to as much as possible. If that includes trade-offs, I am totally willing to take that position. In this debate, I take the idealist position. I do not believe that those trade-offs increase net harm. In fact, I believe in the opposite. A lot of violence that is perpetrated in the real-world stems from a lot of the content that is being demonstrated online. So, while I do acknowledge the negative aspects of this solution, I believe the positives outweigh the negatives. Another example I can give you is pornography. Meta AI, according to Global News, has generated roughly six thousand pornographic images on its platform. That, in my belief, will have much worse ramifications for our society, compared to a certain age demographic having little to no access to an app.

You're asserting that online content causes real-world violence without evidence. That's your weakest move. Studies show correlation isn't causation—countries with less internet access have comparable violence rates. Your Meta example proves regulation already exists and failed, yet you want more of what didn't work. If regulators couldn't stop six thousand images, why trust them to stop the next harm? What specific regulatory mechanism stops bad output that Meta's own systems couldn't catch?