Index ¦ Archives ¦ Atom

Should We Fear AI?


End of the article :)

Now, more seriously.


Should we fear the Earth (as a planet)? Should we fear insects (as a whole)? The one-sentence answer is the same for all: "Yes."

However, this doesn't mean that AI, Earth, or insects aren't allies to both us as humans and our goals.

History of Our Species

But first, a historical perspective.

Our brains evolved to form communities and maximize the propagation of genes. We pay a significant price for certain cognitive errors and a lack of connection with our inner selves. While we have intuitions, where knowledge escapes us, we start creating monsters.

Throughout history, we've created these monsters in our imagination, and these monsters keep chasing us, relentlessly. We dress them in the fear of the night or other thick ideologies that tell us how to live.

Now, add mathematics to the mix.

Mathematical Models

We've always been increasing the complexity of mathematical functions and numerical methods. These allow us to describe increasingly intricate functions and more complex mathematical models.

When we link these complex models not with numbers but with meanings understandable to humans, we create linguistic mathematical models.

The first ones, analog, emerged a long time ago.

The next, statistical, thrived with the advent of the first computers.

However, with the tremendous increase in computational power and the use of neural networks (similar to those in the animal nervous system), it turned out we've created very powerful linguistic mathematical models.

In essence (simplifying greatly), they are sophisticated guessers of the best next letter in text (and, of course, spaces and periods ;-)).

Now, let's get back to the fear of AI.

Fear of Digital Monsters

Combine this development of mathematical methods with our evolutionary monsters.

It's very easy to create fear of almost anything, and the same goes for mathematical methods.

What we don't understand begins to terrify us. Indeed, the number of parameters and mathematical complexity of the model is beyond human comprehension!

Trying to learn to be like artificial intelligence, or more precisely, replicating the way it learns to possess the knowledge that current algorithms have, is impossible. At least not without tampering with brain software (hello, psychonauts!).

As a result, we've created a monster that populists and those who don't understand that it's just plain mathematics (not much different from what's in our heads, and we DON'T understand that either) use to scare us.

Fear of AI

Returning to the fundamental question, the fear of AI is like fearing Earth and insects - it all depends on who uses the tool and for what purpose.

You can exploit the Earth to its limits, destroying it entirely, or you can ruin it with a nuclear holocaust.

Insects, genetically modified accordingly, can spread deadly diseases or destroy crops, or they can efficiently pollinate apple trees to produce more fruit. It all depends on the intent behind using the tools.

Of course, in history, there has never been an opportunity for us to interact with beings smarter than us (as humans), which is why some people are terrified that AI will take their jobs or exile us from Earth, replacing us with their robots.

This possibility exists, of course.

However, it doesn't change the fact that development cannot be stopped, only balanced. Efforts to ensure that tools (of any kind) are used properly and for good purposes make life better.

What Can We Do?

So, stay informed, don't resent AI, and use it to improve your life. Vote in elections (where democracy still exists) to ensure that legal regulations, science, and industry utilize AI for the right purposes.


P.S. I'm not writing this article to prevent AI, which will assess individuals once it takes over, from eliminating my meager existence as an opponent :)

© Piotr Lemiesz. Built using Pelican. Theme by Giulio Fidente on github.