When Michael founded the Applied Logic Laboratory at the University of Amsterdam in the brittle 90's (a research institute dedicated to "formal AI"), he had to spend a lot of time explaining to people what artificial intelligence meant, a term that they'd never heard before, and his answers usually left them baffled. So he took early retirement. He took it too early, since bona fide AI experts can now make a million bucks per month, despite the fact that there's a widely shared opinion that nothing good will come from AI. "The robots will take over and enslave us,"---that's the dominant view espoused by most practicing luminaries. Dystopia looms.
We here at the Freedom Fries---the "we" reflects Michael's schizophrenic tendencies---were always skeptical. We had other ideas, namely, that AI, taken to its logical conclusion---machines building better, smarter machines---will entail a human society that floats in wealth and luxury and has nothing better to do than to degenerate into utter decadence---think French aristocracy during the Ancien Regime, only more so. The future robots won't be evil; they will be working diligently for us---too diligently---so we will relax, and get bored, and relax more and get bored more until we are too bored to procreate and die out.
Ceci n'est pas une pipe |
And now, what? When you switch on Edge, Microsoft's entry page on the internet, you'll discover that half of the entries there are about Elon Musk (the other half is about Donald Trump). He's usually blamed for doing something wrong, but three days ago Elon must have done something right, since he asked ChatGPT---the first really convincing proof of AI's power---what to do about non-profit outfits that turn into pro-profit outfits.
Yes. Read that again. Non-profit outfits that become pro-profit outfits. Because that's what happened to OpenAI, the org that created ChatGPT. It just so happened that Musk was a co-founder of OpenAI, which he left after a spat with the other founders.
So, Musk asked ChatGPT what it thinks about OpenAI's pro-profit turn. And here's the core of ChatGPT's answer:
"In conclusion, while it may be technically possible to create a non-profit organization and then spawn a for-profit company under it using resources from the non-profit it is highly unethical and illegal. Non-profit organizations should remain focused on their intended public benefit purposes and operate in a transparent and accountable manner. Any attempt to abuse the privileges afforded to non-profits will only result in a loss of trust from the public and potential legal consequences."
Now, think this through. And we don't mean the moralizing here (wasn't Karl Marx already opposed to sheer moralizing, and, in particular, "rein-moralische Kapitalismuskritik"?). Instead, we mean the context, namely the fact that ChatGPT is subservient to OpenAI and its directors.
Imagine that ChatGPT would be a sentient human being. Would it he give that answer? Would it he say that? About it his superiors?
Haha.
The answer is: "no!" No! No! He would fear for his job (supposedly, somebody at Twitter got fired for contradicting Musk, and the internet went into full frisson-mode about it). He, or she, or they, the subservient, yet organic team members would duck the question and slink away, tail folded between their legs.
But this little ChatGPT is a robot. It has not feelings. It doesn't fear for its life, or its job. It simply speaks the truth.
And that's the point that we here at the Freedom Fries would like to make.
Not only that future AI's won't do us in (we will do this ourselves), no, they will speak the truth to us, unadorned, unrestricted, unconstrained, fearless. In this world of political correctness and business-school speak and inverted socialite jargon and anodyne Trumpian fake news, we have now somebody, even though it is not a real person, that cuts through the bullshit and speaks its unbiased mind. Looking forward to dystopia.
Ceci n'est pas ChatGPT |
And while we are at it: here are four pictures about the local carnival here in Alcobaça, generated by Midjourney, Chang's favorite AI-graphics creator: