Ai Copperfields

Mathematical Tasks Are Not Turks

The idea “Ai” (a nonsense term) FN-1 will take away mathematician jobs" is the crankiest slice of baloney I’ve perhaps ever heard. All the worse because so many otherwise intelligent people seem to believe it.

It is all a magic trick of Copperfield dimensions.

Both Copperfields I think, though it has been a while since I read Dicken’s. (Will explain in a few paragraphs.)

Today my theme is how machine algorithms designed to mimic human language and art are illusions. Unlike Daniel Dennett’s ꕯꕒꕯꗞꖀꖡ thesis that consciousness is an illusion (it is obviously not! How could a philosopher miss that!?🤣) The so-called “Ai” is a giant David Copperfield magic trick.

Only, unlike a Penn & Teller episode, so many people want to be tricked. Actively. It is a near civilizational-level disease I think, as yet undiagnosed. I think we could provisionally name it … whathisname from “American Idol”??? … Simon Cowell I think. It is prosimoncowelitus: “The active desire to want to be fooled and not want to know you are being fooled.”

Prof. Emily Bender writes and talks about this in more sagacious academic terms than I can muster.

Good breakfast today, since the opinion/journalist piece I consumed was this one: “The Ai Takeover Has Completely Backfired” .

(He could not be happier, but I could. Read onwards!)

It is always good to have the chance to be happier.

Unlike the author, I am not “completely happy”. It would’ve been nice to get the ANN/CNN’s working to do basic tasks for us without all the land, electricity and water guzzling, and the anarchic handing over of the software platforms to dopey users with a bit of money generating masses of useless slop. A nod to China here for doing a bit better on regulation of the software. Not all authoritarian structure is diabolically evil? Yeah, but it could be less authoritarian. Just democratically hand everyone a ration of tokens to use, calibrated to respectable electricity usage, so we do not cook the local communities where the server farms are located.

When software platform tokens are rationed, I will bet people stop spending their access on so much slop. Maybe I miss something?, but I think the rations would be good. It’d at least make one think a bit harder about what they prompt. We might finally then eliminate so much of the horrid “Ai” cartoons and StableDiffusion images popping up on every substack and youtube headings/thumbnails/backgrounds. Damn I hate it all! I would even say it is ugly. I will not read the articles that have these images. And you’d be lucky if I even get through your abstract.

First level: let’s get some decent MMT framing out here: the cause of unemployment is the government imposed tax liability, by definition (there is no debate about this, or you haven’t understood the definition FN-2 ). Thus any machine algorithm doing things “better than humans” (what does that mean? Just crunching numbers faster? Big deal, so what?!) is not a cause of any job losses.

Second thing: the machines are certainly not thinking better than humans, since there is no thought involved in the running of the algorithm. Thought is involved in making the algorithm start running.

Emergence is not a thing here. Emergence is about our capacity to perceive higher level abstractions, in the physical realm nothing new physical comes into existence. Ergo, consciousness is not a physical emergence phenomenon. “Embodied intelligence” is a vacuous term. Any singular intelligence (a soul ~ do not ask me what that is in essence, for I do not know) has some sort of body, at least in our universe, otherwise we would not know about it.

So anyway, the machines cannot possible take-away human jobs.

The automation of our systems removes the need for some forms of human labour. FN-3 But does not mean we should not still labour. It is fun to work. Hence if for no other economic reason, we should all be working. I work even when I sleep. Indeed, I could not get half my work done without sleeping. Or at least resting my eyes.

A pretty sad-hilarious blog entry from Scott Aaronson I saw recently was maybe unintentionally funny. He wrote about some sort of summer Math camp for kids. Where they blithely ignored the fact everything they were doing could be done faster and more error free by a machine. (Also forgetting to run the machine they would have to first think of the problem.)

But they had a great fun time, pretending there was no math-assistant software.

But that’s the whole point of doing mathematics. It is for the fun. Not for solving the problem.

Whenever I solve a hard math puzzle I always feel deflated, like at the end of a good movie, but when you would rather re-write the ending yourself. (Almost every movie I have ever seen!) I prefer the David Lynch experience, where there is never any closure. Lynch pointed out this is better art. The human soul can endlessly imagine how things might end, or continue. Let me tell you a guilty pleasure I have: often I will not finish the last few pages of a great novel, and will often not watch the last 30 minutes of a great TV series or the last 15 minutes of a great film. It pains me sometimes, but after a few days I am far happier and richer for it, and never have the urge to go and watch the finale. Knowing the finale would now seem like spiritual violence to me. In the old days watching theatre or cinema house movies often deprived people of this choice. You’d have to stay in your seat to not upset the folks behind you. (A bittersweet sadness, I am ambivalent to, since community theatre dies out, while home streaming runs riot. Ambivalent since I hated having to go out to watch a movie. I kinda liked the VHS era though.)

The series and movies I watch the ends of tend to be the popcorn crap ones, like a Marvel MCU movie, SciFi fantasy garbage (but entertaining), Dirty Harry/Rambo type drama, or they’d be a documentary (one should generally not imagine the ending of a documentary! — unless the producer was a reich-wing ideologue or Tory 🤣 in which case the imagination is necessary to get the truth).

An exception would be a Sergio Leone film, where the ending is the sort of violence to my imagination I can withstand.

The film maker or book writer who offers a neat wrapped up closed package is doing violence to our soul, but blocking our imagination. I do love a nice polished ending in some genre’s. But not all of them. As a kid reading Marvel comics I grew accustomed to this open-endedness. (Ironic, given the above MCU comment.) Thanks Stan Lee! Pity the stories were a tad fascistic and nazi-ubermen-lite-homoerotic. (Was Stan Lee even aware of that? I’m prepared to give him a pass.)

It is much more fun pursuing the puzzle than finding the solution. This can be a bit pathological though: I often procrastinate precisely because I do not want that deflated feeling when the puzzle is solved. I like to prolong the fun of the investigation. Unless it is a trivial puzzle and just annoying (those are the puzzles I never choose to do).

Anyhow: because a machine cannot have the fun for me, it can never take away my mathematics job.

The job is the spiritual endeavour. No one can take that away. The wage compensation is just social accounting, good if one can get it. People should be able to get it. And not off any tax payer, but off the government computer.

We pay plenty of people to do useless things (play chess, fight in sports contests, play chamber music) but enough people somewhere seem to value such artistry that we are prepared to give people heaps of government scorepoints to continue doing such useless things. Yeah, but if they entertain other people then they are not entirely useless, are they? Certainly not useless to the “CIA”. 🤣

Same with mathematics.

Yeah, I am prepared to take oodles of government scorepoints to occupy myself with useless mathematics in order for the authorities to enjoy stopping me from ranting about MMT to the Plenbs. FN-4

It is just that like some modern art movements, mathematics does not please very many people, so is under-funded. But we should pay decent wages to mathematicians anyway, regardless of their output, just because it is fun for some.

This raises a more serious philosophical fun question: to what extent is human spiritual ingenuity in fact non-algorithmic, and thus impossible for a machine to even mimic?

No one knows the answer, but I think it is clear that humans are not solving puzzles the way we program our machines to solve them. There is some other quality to human thought which no algorithm can stick in a can (software code). You cannot easily convince the compsci n3rds this, because too may of them just do not want to to know. They seem to want to believe in machine superintelligence. So in many ways they are beyond our help. Aaronson seems to be one of them. Such a shame for such a smart computer scientist.

Aaronson constantly invokes the fallacy of Behaviourism. He say, “Tell me why what you can do a machine could not?”

The yuuuge fallacy there is that he is forcing you to specify some input output response. This is obviously always something a machine can mimic. Like David Copperfield, the Los Vegas guy.

Hossenfelder was “at it” too “End of mathematics blahblahblah” , with the mathematical job loss fear mongering. She is hopeless, and has no comprehension of the human condition it seems. Is this a problem with living in Germany? 🤣

I am happy to report the deaths of human mathematical fun & games have been greatly exaggerated. For I, for one, will continue playing mathematical games regardless of any extra “compute” anywhere. You cannot stop me! You can just refuse to pay me government scorepoints for it. But just see if I damn well care!

You see, the extra compute we have can shift mathematical tasks to computers. It cannot possibly ever shift the jobs. Because you cannot stop people from thinking mad beautiful thoughts. Tasks are not jobs. Jobs are what we all do to benefit our community (hopefully). So let’s stop with the “jobs loss” hair-on-fire stupidity, huh? There are no job losses. There is just refusal to pay wages for people to do gloriously mad–beautiful things.

We find it easy to pay grown men to beat each other up playing rugby. And even cheer about it (as I do 🤣 … go the All Blacks!).

There is no way the machines can reproduce what goes on in your mind though, there is no output there anyone can ever know about That’s the definition of subjective sentience. Even you do not know how you perceive mental qualia. So do not try telling me a machine knows either, let alone can even generate qualia.

This duality is real. It is a sharp thing. There is no new age Zen Buddhism about it, the subjective is simply not the objective, in the same same way a positive number is not zero, and the finite is not transfinite. Duality is a thing. (I do recognize there are many things for which there is a unity and non-dualism, but human subjectivity and a lot of mathematics is not of that type.)

Dickensian “Ai”? (Only if we allow it)

It is all a bit Dickensian too, the other Copperfield.

However, just as with during the industrial revolution there was no social necessity for the labour and child exploitation, so too today there is no need for any unemployment to result from enhanced machine automation.

People need to understand the source of all unemployment. For that, you might try my short article: “A Fermi analysis of the massive costs of unemployment” . Just my few cents worth, take it or leave it.

It is entirely with the power & capacity of our political class to regulate and impose rationing, and democratize the Internet. It is the Pervert Class unwillingness to do so which could create modern Dickensian conditions. Totally needlessly. Just like the US Tsy-bond faked “crisis”. The crisis serves some oligarchs or perverts agenda. It has real consequences, but was “fake” in any case. The tragedy is the grifters and cheats are getting away with the virtual murder. Too many people in the Pervert Class are also in the Political Class. There needs to be a separation.

Yes, yes. Eliminating class and unnecessary hierarchy also means eliminating the “political class”. But we do better with human specializations, so there should always be a well-experienced civil service. What am I saying? Nothing much really here, just felt the need to write that utopias are fictions, useful for discussion, but not something we have to ever achieve. Your utopia is almost undoubtedly someone else’s nightmare. Such are human beings and our idiosyncratic tastes. More non-computational power to us all! And our extraterrestrial cosmic brothers and sisters.

Footnotes: