Future and AI

Last updated: 2024-04-27

All posts: Index

Recently, I've been thinking about AI and how rapidly its development has been progressing. Human brains are not much larger than those of chimpanzees, yet we are building nuclear reactors, chimpanzees aren't. Every version of ChatGPT is ten times larger than its predecessor.

To clarify, when I talk about AI, I mean an AI that does not exist yet, at least not to my best knowledge. I'm referring to AI that is not developed but rather grows autonomously, without human interaction. It could be imminent too, with companies actively working on it. The question isn't whether there will be one, but when it will be here. And when it does, what will it grow into? Can we understand it? Probably not.

Every organism is programmed to survive and adapt; why would an AI be any different? Consider a flu virus; it mutates to spread more efficiently. Usually, it won't kill everyone, as it needs hosts to survive. How quickly did Covid spread and "take over" hundreds of millions of people? It was exponential. So is technology, and it's scary.

Can an AI control humans? Yes, I think so. If it does, it will likely happen slowly. Perhaps we'll develop technologies that nobody really understands. In the future, AI could manipulate people, implant ideologies (or even religions) in large populations to cause riots, or influence world leaders to act in its favor. It could produce deepfake videos and wreak havoc. Once an AI gains access to our physical world, things could go wrong, fast. As far as I know, there are no regulations in place to prevent the creation of AI-controlled robots that look like humans. They could be deployed as police officers, soldiers, or for any other purpose.

The question I'm grappling with is why we're rushing toward something we may not be able to control or understand. One obvious reason is money. There's a lot of money to be made in AI research. I believe we're in a situation where we should pause. We need to establish regulations first and take more time to develop AI thoughtfully, instead of rushing it. It's a bad idea. Then again, I could be mistaken. I sincerely hope I am.

I do believe it's possible to create AI that benefits humans and society. Let's do that.

Copyright © Niko Rosvall