The One Thing We Should Be Afraid Of With AI: Undermining Democracy

I'm not prone to doomsday prophecies. Most criticism of AI is overblown, but one danger that does need more attention is the role AI will play in exponentially spreading misinformation and undermining democracy itself by making it impossible to separate fact from fiction.

An article on Gizmodo described it as the "clogocracy", where language models are used to personalize and weaponize media to change voting behavior and polarize citizens even further. In this scenario, the candidate with the most unscrupulous and manipulative AI machine is the one that will win. So how do we prevent this from happening? Better privacy protection, for one thing. Media literacy training is another. As the article notes: "every bit of that information that companies or policymakers deny the machine would make it less effective." Other proposed solutions include requiring disclaimers for content generated by AI or banning so-called "replicant speech."

The one thing that seems certain is that the 2024 US Presidential election will present a global test case in algorithmically accelerated manipulation versus our individual abilities to seek out and find the real truth for ourselves amidst the lies.