Artificial Insanity

It now has been over a year since the current AI trend started. The name "artificial intelligence" is of course stupid as is the senseless hype by investors, but the technology itself and what it enables is of course interesting. I wanted to take some time to experience it more and think about it, but I believe enough time has passed now.
There are without doubt nice things that can be done with it. Text and object recognition or automatically improvement of images are such examples. There are some problematic things like the enormous electricity consumption of AI. But mainly there are three big problems, that sadly easily outweigh all the benefits for me.

Thoughtless Creation

AI makes it easy to create things without understanding how and why it's actually done. That is part of its appeal for many people. You just type in a few prompts and you get your images, texts, code or (possible made-up) information.
The process and knowledge how the end result is created and the context is really important though. How does the image get composed? Why is this text written this way? Why does the code work how it does and what are the possible pitfalls? Where is that information from and was is its context?
This mindless creation ignores all those things, that are really important for quality work.

Power Theft

A few rich people and companies just scan content to help train their AI. There basically never is informed consent. They just profit of millions of people's work without giving anything back. Just collecting more money and power for themself.
If an ordinary persons just downloads a song to listen to it without paying or consent from the creator, it's illegal. But if it's done millions of times by organisations and people with billions in the bank to mix it all together, completely control its use and make enormous amounts of money from it, we should all just be okay with it? Of course we shouldn't.

Garbage Floods

Trustworthy, manageable communication is important for our daily life and our democracy. The internet has made this easier and more inclusive than ever, but AI has the power to destroy that new openness.
The problem is that AI makes it easy to create content, that can be easily varied and at least on first sight often can pass as human-created. So it makes it easy for bad actors to flood and overwhelm places where human dialog happens without being spotted easily.
In it's most basic form that could just be denial-of-service attacks on for example customer support, but it can get quickly more dangerous from there. You can make a place basically unusable by overwhelming it with fake stuff. AI could be used to steer debates on social media with fake accounts. It could be used to create an avalanche of fake media of an event. It could be used to created big amounts of fake calls or e-mails to elected officials influencing their decisions. And that is just the stuff, that immediately come to my mind. I'm sure there are far more ways to create havoc.
Of course all those things can be faked already, but it requires a lot more effort for far more easily spottable fakes than AI will. I don't think most people really understand the extent of what could become possible yet.

Conclusion

Those three big downsides seem far worse than the positive things AI enables. Some of them can still be prevented while I fear other probably will inevitable happen.
This is another important reminder that such important decisions about what technology we use in which ways shouldn't be made only by a few unelected, random persons.