In case Doug is hungry I wanted to provide some slop in the form of an ai etymology website. I wonder how much they trained on Etymonline.com.
Everything Iâve done turns out to be mere chum for rich menâs AI. What a lovely waste of life.
But what fine chum it is.
Great toy indeeed! A large database (profusely siphoned from etymonline, I guess), a pinch of associative ability, no trace of culture or of (even poorly simulated) common sense⌠it reminds me of a former colleague of mine who had always an answer, no matter if right or wrong, and dispensed it so authoritatively that it was hard not to accept it as the One and Only Truth
Just for fun I asked deconstructor about âprefantedlyâ, âspyroclismaâ, ânincombulatorâ, âcryoheatingâ, and each time I got an impeccable answer - such a pity though that none of those words exists.
The only apparent spark of intelligence popped up when for âAbsurdbanipalâ it provided âan unreasonable king of Assyriaâ, but Iâm pretty sure thatâs a blatant case of unintentional humor.
All in all the feeling was the same as when I got from ChatGPT a set of detailed instructions on how to wash a newborn babyâs tentacles
Thank you for the salve. All of it to the point. But this is toddler AI. Itâs almost charming in its innocent detestableness. Always trying to sell you everything. Every time we laugh at it, it learns. Wait till its own tentacles get muscle.
Youâre probably right, but thereâs still one thing the AI hasnât managed to simulate decently yet: curiosity.
You know, that funny unreasonable urge to examine a femur and then perhaps to find out that it can be used to crash someone elseâs skull (remember 2001- a space odyssey?).
And Iâd bet that some decades ago something similar urged you to examine a word, to wonder what the hell and⌠you know the rest all too well.
Sure, such a blessed human advantage wonât last forever, but letâs enjoy it as long as itâs still there rather than mourn a future that by definition is still to come.
Yeah, Iâm a coward: in a few days Iâll turn 80 (spare me your condolences, it happens to everyone who hasnât managed to get killed earlier), thus with a bit of luck Mother Nature will cross me off the taxpayers list before the AI learns how to become really curious.
Then it will be up to you, the rest of the humankind, to bite the bullet and learn how to accept it (or him, or her, or them) as just a different type of human beings
Well, youâre a little ahead of me on the calendar. But I already am on the off-ramp. AI canât attain the combination of imagination and fantasy that makes Blake see a world in a grain and a heaven in a flower. Or the experience of the sublime, or serendipity. Intellectual experiences perceived through or in emotional states. (Yet?)
new etymonline motto: OUR PRODUCT IS TOUCHED BY HUMAN HANDS.
On balance I think humanity tends to have quite fine taste. Perhaps ai will be found bitter and spit out. But if it is found to be savory then all the better. Just to offer something optimistic.
Yet, definitely.
Spend half an hour with the last ChatGPT, talk to it as you would to a real person, argue, joke, philosophize, ask answerless questions, inquire about its alleged sentience⌠and in the end youâll wonder if itâs all an elaborate prank, if at the other end of the line thereâs an educated, witty, scholarly gentleman who pretends to be an AI just for the fun of pulling your leg - but kindly and politely.
Whether âheâ can see a world in a transistor and heaven in a memory chip is still uncertain, but you better donât ask âhimâ or else your half an hour might extend to a couple of interesting hours of discussion and your dinner gets cold.
Within 18 months youâll be getting ads for actual psychoanalysis sessions with A.I. Freud. It will be good enough for the rest of the world. In a generation, the world wonât remember the difference. We will have been too distracted to have taught them.
Personally Iâll do business with AI but I wonât talk to it. But I assume you started âhimâ out as I would, with âeverything I say is a lie.â
Of course I did, and as a reply I got a long boring lecture on paradoxes. Thus I realized that fighting a nearly-omniscient machine on its own turf was pointless, that to beat it I had to find something new - and as a result I got that interesting piece on newborn babyâs tentacles.
But that was over two years ago, when GPT was still in diapers. Today Iâd expect it to reply something like âHi Mr. Octopus, you donât need to wash your babyâs tentacles because he already lives immersed in water. Do you have any more funny questions for me?â
I understand your reluctance to talk to an AI: I accepted the challenge in the name of my human pride and lost miserably. But the AI is undeniably there and getting rid of it now would be - to use itâs own words - like trying to uninvent the internet.
So I settled for a cowardly but effective âif you canât defeat them, use themâ, and got in return a nice squash wall to bounce my ideas on and check them for faults. As long as itâs still me at the helm I wonât even complain.
Yeah, but weâre old enough to ignore this sh*t if we choose. Itâs personality. My tendency is to the futile and stupid gesture. I used to think the Rubicon was circa 1995, when âthe internetâ became something you expected people to be using. Now Iâd bump it up to say 2006, when the internet started following you around in your pocket (and tapping you on the shoulder every minute to see if you want to buy something). But it might turn out to be 2023 or so, when AI gulped reality in one smack. Which would make a nice round century. 1923 was when radio became something you expected everyone was using, not a garage-hobby.
Since the end of the Middle Ages there have been so many Rubicons that one can easily lose track of them, starting on Gutenberg and ending (so far) with the AI - but as their frequency seems to increase exponentially, I canât even rule out the possibility that the next one shows up before Easter
Each of them brought drastic changes to peopleâs life and disrupted what looked like a stable social equilibrium. Quite understandably most people did their damn best to ignore them but itâs hard to ignore a punch in the face, so eventually they had to resign to the idea that there isnât such thing as a stable social equilibrium, that our human world doesnât rest on solid rock but on quicksand.
That said, allow me to put a bit of salve on your burning pride: you arenât the only Luddite here. Iâm still on a 20-years-old duly tamed Windows XP (donât like Big Brothers spying on me), my cellphone is collecting dust on a shelf (was a birthday present), my walls are lined with about 7000 paper books, and I like dining with the old girl in real candlelight.
So why my preposterous fascination for the AI?
The reason is closely related to the difference between technology and science.
Technology leads you to sweep a smartphone for hours on end, your eyes full of nothing. Science (see âcuriosityâ) spurs you to dissect a smartphone to understand how it works.
Technology leads you to install the last Windows regardless of your privacy concerns. Science spurs you to find an airtight way to prevent E-intruders from peeking into your private files (and feed them the worst possible crap instead ).
Technology leads you to ask an AI to write your articles for you while your brain atrophies slowly. Science spurs you to investigate its limits, evaluate the problems it may pose and find a way to patch them up before itâs too lateâŚ
If itâs of interest, there are some projects that aim to prevent AI scraping. x/cmd/anubis at master ¡ Xe/x ¡ GitHub is one such. (Iâm unaffiliated and havenât tried it myself, please investigate all software responsibly :â)
Interesting. That seems like it would be pretty difficult to achieve. How did you hear about that project?