Mx. Chara Aznable (they/them) of Pnictogen

RE: https://toot.community/@fak/116353999914043336

I suppose that I can justly be called an #LLM "hater", because I have nothing good to say about that particularly manifestation of technology. Here's a reason for my "hatred", which I confess does not seem so much like hatred to me, but rather intellectual disdain: people are encouraged to use them for things that one should not ever expect them to be good at, not if their true nature as technological devices were accepted matter-of-factly.

Once again I am forced to explain that #LLMs construct their outputs stochastically: given some input and asked to construct a response, the device is choosing words and phrases randomly (well, pseudo-randomly), that is to say stochastically, according to the statistical likelihood that someone, some human being mostly likely, has in the past emitted text in response to the words and phrases in the given query.

The greater the range of permissible randomness (or pseudo-randomness) in the LLM's response, the more "creative" the LLM is said to be.

Why would you expect such a device to be good at telling the time?

9:00pmWeb
Show threadMx. Chara Aznable (they/them) of Pnictogen14h ago

By "pseudo-randomness" I mean the usage of deterministic algorithms which are thought, on the short term anyway, to produce a stream of output that's indistinguishable from truly #random noise generated by some physical process. There are numerous methods in #programming for generating what looks like uniformly distributed "random" numbers, but which are in fact generating a predictable sequence with an extremely long period, such that using the pseudo-random numbers generated by such a technique is thought to be indistinguishable from physically generated randomness.

Does that all sound just a bit uncertain and hand-wavy? It should. The question of what "randomness" really is, and how one can gauge the quality of pseudo-randomness vs. the real thing generated by a physical process—these are not settled questions, and thus one should immediately be on guard against any claims made by overeager #computing people who want to act as though this is all a boring settled problem, something out of a textbook. That sort of easily bored computer geek just wants to throw prefab algorithms together on a server bank and hopefully make money thereby.

Show threadMx. Chara Aznable (they/them) of Pnictogen14h ago

One can at least say with tolerable certainty that there are exceptionally bad methods of pseudo-random number generation, methods so bad that the resulting numbers, while still looking "random" at a naïve glance or passing basic frequency tests, are actually very specifically distributed throughout the space of all possible values. "Pseudo-randomness" can in fact conceal a great deal of predictability.

I'd very much like to know how big of a problem this is, in #programming. And here's where I'll be blunt: thanks to capitalism and broken methods of U.S. #education, I do not trust the average computer programmer to have thought about such problems at all. Corporation and corporate #tech companies generally don't want computer professionals who worry over fundamental weaknesses and pitfalls of their field; they want people whom they can plug into a management hierarchy, people programmed to do a limited range of expected duties.

Show threadMx. Chara Aznable (they/them) of Pnictogen13h ago

Programmers have been acculturated in such as a way as to encourage a kind of slack-jawed amazement at the purported limitlessness of what their machines can achieve, and thus they have fled from acknowledging that modern #computers actually have some serious issues when it comes to representing numbers and doing #mathematics. There's been such a massive overemphasis on pure speed and bulk of computation that issues with accuracy and cumulative error in calculations have been swept aside, assumed to be taken care of with the proper #software—and yet all our software and its development is in the custody of people who are culturally disinclined to want to think about such problems.

Let's be honest here: today's top-flight #programming people have been schooled to fret over some very strange non-issues. They have a very poor sense of priority, which is partly why they've gotten so political lately: the #technology elite seem to think that their machines are already so perfect and wonderful and amazeballs that the real problem with computers is that humanity is too stupid to use them.

Show threadMx. Chara Aznable (they/them) of Pnictogen13h ago

The #tech elites are "optimists", i.e. they have a naïve faith that there'll always be a faster machine, a more powerful machine, a better machine. They're not entirely sure any longer what "better" means, as an abstract principle, but they're fairly certain that corporate entities able to hoard massive piles of venture capital are surely "better" at something. I mean, isn't that why people gamble the most money on something? Because it's the BEST thing?

(coughs)

The overall result has been a #computing and #programming culture that—at least going by what tech ventures get the most funding and press coverage—is weirdly stagnant and conservative and backwards-looking, despite all the noise and hype about being innovative and "accelerationist". Underneath that hype is a deep-seated conviction that one must stick with whatever technical developments have already become entrenched in the industry.

Show threadMx. Chara Aznable (they/them) of Pnictogen13h ago

Underlying problems and defects have been accumulating throughout the #technology sector over the past several decades, but it's been possible for the businessmen and entrepreneurs to pretend there wasn't any real trouble, nothing that couldn't be fixed by the usual means of making all the hardware faster and denser and better. #Software could continue stumbling along in a chaotic and semi-regularized state, written by people who refused to obey basic principles of engineering ethics—ethics which require the engineer to build things for others, not for their own personal satisfaction, as if they were making a solely personal work of art.

And thus we arrive at a central difficulty with computing technology: it's become the domain of people who have a doublethinkful and treacherous attitude towards #art, in all senses of the term. They want to think of themselves as artists...but they also hate artists.

Show threadMx. Chara Aznable (they/them) of Pnictogen13h ago

Are there filmmakers who think that #cinema has somehow replaced other artistic media, so that humanity no longer needs novels or operas or musical albums now that they've got movies?

Probably, yes, such people exist. There probably exist people who think that the cinema isn't just a method of artistic expression, but the only method, the one that's better than all the others. The landscape of #computers has become yet another medium for artistic expression—and it's one specially inclined to inflame feelings of artistic supremacy. The computer person is apt to be tempted away from other media, thinking that such media is now old-fashioned and superannuated and ripe for total replacement by the computer.

Show threadMx. Chara Aznable (they/them) of Pnictogen13h ago

I haven't seen the expression much lately, but I recall how "dead tree" became a sneering dismissal of books and other paper media during the 1990s #Internet boom. I'm sure that the reactionary #tech nerds of the Curtis Yarvin sort still burble about "dead trees", as if their own medium weren't also a dead thing and the product of destructive industries.

In sum, computer culture really ought to learn some humility! There's been a general unwillingness to drop the 1990s-era that computers and the Internet were somehow the ultimate workaround all social problems—as if the Internet had solved the issues of freedom of speech and censorship and public access to power (and so many others) merely by existing. Conversely there's been great reluctance to admit that computers are not a solved problem themselves. The tech itself isn't just imperfect; it's arguably unfinished.

Show thread💡𝚂𝗆𝖺𝗋𝗍𝗆𝖺𝗇 𝙰𝗉𝗉𝗌📱4h ago
@mxchara
"have been schooled to fret over some very strange non-issues" - using goto
Show threadMre. Dartigen [maker mode]10h ago
@mxchara I can't speak to other areas of computing and I don't consider myself to have more than an extremely basic understanding - but I recall that randomness and the difficulty of truly random generation is a major issue for cryptography and security. (And so you get all sorts of attempts at work arounds like the wall of lava lamps, images of which are supposed to be able to introduce some true randomness because the variables behind what a lava lamp looks like at any given time are extremely complicated.)