The Worst Case Scenario for AI Regulation

Most immediate fear is that we're going to have so much misinformation now, nothing we've ever seen before. I'm worried that large language models are in the end going to undermine democracy. It's hard to predict where things will be 10 or 20 years from now. We should require people to label any AI-generated content as such and make it actionable if you don't. The 2024 US election is almost certainly going to be heavily affected by misinformation. If nobody trusts anything, it's hard to have people know how to vote.