We’re Underestimating the AI Threat to Democracy and Humanity

This post originally appeared at Government Cheese – Chronicling the collapse of democracy

The climate crisis is like an asteroid that is heading towards the Earth. AI could be another asteroid coming at us, but much sooner.

I used to think AI was amazing. Then I tried using the stuff they’re putting out there for various things and it was terrible, wrong and annoying. So I started to scoff at it.

I’ve changed my mind again. We shouldn’t disregard it or scoff at it. It is already replacing coders, and it is good at lots and lots of other things. It will get better and it will get better really quickly.

Here’s a prediction: Without regulation AI will be used to manipulate the 2026 elections in ways we don’t see or understand. In other countries the same will happen. The oligarchs are guaranteed to try this, and might well succeed.

The goal of the AI effort is “superintelligence.” This is Artificial General Intelligence (AGI) that can learn and make itself better. Once it gets to that point it will immediately accelerate its own learning and will very, very quickly surpass anything people can do.

The people running the AI companies are hoping to “get there first” and use it to take over the world, for themselves, personally. For their own benefit, and never mind anyone else. I think that’s obvious.

Imagine a tool that lets dictators anticipate and disrupt anything – anything – that might try to stop them, and that can manipulate the public with a power beyond our comprehension. A tool that is infinitely smarter than us or anything humanity has ever encountered. People have been able to overcome people in these situations, but this will be different – if we let it happen. It is the INTENT of Musk and the others for this to happen.

So here is a warning from someone on the inside who understands what’s going on. I listened to the podcast while walking. Look for “GZERO World with Ian Bremmer”

“OpenAI whistleblower Daniel Kokotajlo on superintelligence and existential risk of AI”

Here it is on YouTube. (The settings “gear” icon lets you increase the playback speed.)