Is that intelligence already or can it go away?

What, not another piece about AI?! ChatGPT is making waves, most people have heard about it, but is it really “intelligence”? What is artificial about it and why are so many so enthusiastic about it? Some people are already see in it it the end of the world, but…

An exciting question in this context would be: What distinguishes us from artificial intelligence? It can do everything: answer ANY question, cook, clean, etc. – no, actually ChatGPT can only answer questions. Why are so many people in such an uproar now? It’s apparently already replacing us – as can be read here and here. Ok, calm down, it says here that it also creates jobs. Phew, tricky, what now? Maybe it’s just fear of the unknown.

A few days ago, we were talking about AI again among our colleagues and the question came up: what can’t an AI do better than we can? One answer is: ask questions! But especially intelligent questions! You can try out for yourself what intelligent questions are. Grab a child (preferably your own, if available) and ask them: “Ask me an intelligent question” – you will be surprised what comes out. But what is an intelligent question? The question(s) that many other people ask? ChatGPT could even do that, it’s even a simple analysis. But is it intelligent? Or does an intelligent question take into account the context of the questioner and the answerer. Hang on, then we would have to link ourselves completely and directly with ChatGPT – is that you I hear, Elon?

Hopefully I’ll be dead by the time that happens. But seriously, that would be a task. You can see that this was not the focus during the development of ChatGPT. It can supposedly answer all questions, but what the question is, is our decision! Here perhaps Weizenbaum’s “Eliza” was more intelligent, because the program kept a conversation alive with counter-questions, and that was already possible in 1966. Weizenbaum was far ahead of his time – since then above all a lot of computing power has been added and better abilities in dealing with language, that’s why ChatGPT seems so “real”. But do we really have a cognition problem in the form of “give me an answer, baby”?

So, don’t be afraid of new technology – but do look at it and try to get to the bottom of it! Then maybe it will help you, maybe not. It is up to us to make the best of it.

Original text: MHA
English translation: BCO