This is a complex topic. Of course there are way better informed specialists who can correct me on my guesses and stuff. So to speak: this will not be a technical deep dive into AI and big data. I’m just lazy to read into that stuff myself.

But I feel the urge two throw in my two cents on all this talking about ChatGPT, Bard and other AI tools out there.

From my point of view, born and raised in a time where computers would fail often and you needed to understand and creative to get it working again, AI is all a big scam machine which got the mainstream media + people who are born and raised with tech that “just works” in control. I could see that the new generation is seemingly blind if something is not working as intended and don’t have the urge to check on why or how it works or even cross check information.

And here comes the new marketing tagline “AI” which conveniently makes stuff out of nothing. It calls all these people because it suggest that it is a magic thing, which outputs something cool and seeminly automated. Independently of the right or wrong because these don’t know, or don’t want to know the alternatives to cross check.

So how does it work?

It is all a predictment. There is no “thinking” brain behind. Everything is trained with big data sets. Which was also a marketing tagline a few years ago, remember?

I saw a good interesting video from Ultralativ, sadly in german only. But it rounds up very good without getting into to deep details.

The very interesting stuff for me is not how AI is gonna “change the world we live in” because it outputs something wonderful, as some media prophecies, it is how we are gonna live with long time services which relys on data collecting. The first consequences can already be seen:

• Reddit restricts the API; third party apps are not working anymore;
• Twitter is slowly going behind a paywall, not only on the API;

These services have a big chunk of user data and of course the traffic and ressources managing those are no joke. Then imagine a big “AI” is coming and scrapping all these data for internal use and training. Then it gets hyped and now there are coming many alternative models which also need training data and scrapping on every corner.

On the long term there will be a lot less free services around because everyone collects data for the “next big thing” which is only useful for creating marking texts; manipulate pictures or creating human sounding scam E-Mails for even more spam in the world wide web.