The future of AI is not the Terminator or the Matrix

At Cyber Protection Magazine we are trying to figure out how to incorporate #AI into our process. It isn't easy. One option is to use it to post all the press releases and BS studies and surveys we get. That boosts our SEO so we get more reach, but it also means stuffing our site with mediocrity.

Today, our marketing VP was suggesting we use a new AI tool to handle all our #SEO since every SEO and #leadgen service who cold calls us seems to know nothing about the #publishing industry, even when they claim to know it. If SEO is a necessity in today's world, might as well go with the mediocrity that takes the least work and costs next to nothing, right? At least, that seems to be the biggest selling point of AI.

That got me thinking about the predicted AI #singularity that everyone fears, when the machines take over the world and humanity. The two most popular themes are the Terminator and Matrix series. But after seeing the trajectory of adoption for social media and now AI I think there are two movies that have a better handle on the actual singularity: Wall-E and Idiocracy.

To save you from having to actually click through to those descriptions, both movies describe a future where humanity has turned over the earth to billionaire oligarchs and corporations, forcing humanity to the point of extinction leaving robots to organize and pile the detritus for the rest of eternity. In the former, a multinational conglomerate run by a self-assured billionaire creates a luxury space travel service for the wealthy to let them get fat and lazy while the robots try to heal the earth. That one is a bit on the nose right now, although we have elected a professional wrestling fan with a strong dose of misogyny as president again, so the second movie isn't too off center.

Recently, I read a story about how AI is copying everything and producing content that is so much worse than the mediocre stuff we see on the internet. The reason is pretty obvious, if you accept Sturgeon's Law (and most people do). Since the most popular AI engines - OpenAI , Perplexity , etc. - are trained on everything that was on the internet up to a few years ago, that means 90 percent of the training was done on crap content. You know it, they know it, and anyone who doesn't know it suspects that is true.

On this week's episode of The AI Fix, the hosts did an experiment with OpenAI asking whether 9.11 is larger than 9.9. It got it wrong. So I did the same experiment, got the same wrong answer so I asked for it's reasoning. It said, "Because 911 is 812 larger than 99." Apparently the algorithm has a hard time recognizing decimal points. That's problematic if you are doing math with an AI.

There are AIs that are not trained on all that crap. They are fed targeted and vetted content for very specific purposes. The problem is the vetting isn't all that great. A fairly recent study in the UK found that AIs trained for medical diagnoses are not that great, either, basically because the training process was so flawed.

Earlier this year, I asked Bruce Schneier for a prediction on AI. Schneier is very positive on the future of AI and said once the coders got hold of it, the hallucinations and inaccuracies would be a thing of the past. However, a few months after he said that, computer science executives and coders started getting laid off and leaving the major AI companies en masse because the companies really didn't appreciate their input. Slowing down development and all that, don't you know.

As I cogitated all this I saw that the drive toward mediocrity, and everything below, was the eventual future of AI and its impact on humanity. In politics, media, technology, academia and every human effort, we are demanding to turn our backs on facts and reality in place of misinformation, rumor, and superstition.

The future of AI is not domination of humanity. It is in the democratization of stupidity.