Shut Up About the Water
We don't deserve to claim to care
Merry Christmas! This post is overflowing with the opposite of the Christmas Spirit. This will be the last ANGERY GRRR post from ol’ PGB in some time. I have joy and delight to offer this beautiful world! I promise to share it. For now: oh lord I’m mad.
Some internet-famous nerd who has profited greatly from his direct and eager contributions to ruining the fucking world recently shared some words about how bad AI is because of water tables and natural resources and the like. Other not-so-famous internet nerds chimed in on various websites where nerds chime in on things, largely in agreement!
Sorry: I just do not believe you. I see no courage in your convictions.
It’s not that there’s no harm being done here. Of course there is, and of course it’s bad! It’s a legitimate point of contention and a legitimate discussion. These technologies are wasteful, and they are causing harm, and one way or another we must reckon and deal with this! The problem, dear reader, is how incredibly hollow it rings coming from the very architects of some of the worst aspects of modern society.
You guys1 do understand that we — as an entire professional class — have negative credibility on this issue, right?
“Oh, the water tables, you see!” bemoans the man who’s devoted his career to optimizing database tables which store that most private of information which they have no right to store. “The electricity bills!” protests the SRE responsible for ensuring server fleets ten-thousand strong successfully deliver mind-poison; such pride she takes in the reliability of this malevolent infrastructure. “Our beautiful planet’s natural resources!” cries out one of eight thousand product managers for yet another goddamn social media website which is distorting and warping and, yes — in the words of Rob Pike — raping the once-beautiful minds of its users. “How could these AI companies do this?”
They’re doing it the same way tech companies have ruined every-fucking-thing else: with the eager, gleeful, never-questioning support and diligent work of expert engineers like the esteemed Mr. Pike.2
I should explain why I’m so A N G E R Y over someone making an argument I completely agree with: it’s about revealed value systems.
Many tech companies, and their constituent well-paid busy worker bees, have been feverishly working for decades now to strip out all human connection and meaning from our world. Every possible point where genuine togetherness rears its ugly head has had oil derricks plopped on top, turning once-beautiful humanity into wells of profit. There’s oil in them thar minds!
They have worked so hard and for so long at this task: put a screen and ads and also a paywall between all humans and in front of all beauty. It’s the only thing they know how to do! Commodify; extract; rent-seek; ruin; do all this to everything which contains true value. Rinse and repeat and repeat and repeat and don’t even bother with the rinsing anymore because that’s not efficient, you see, and that company got away with not rinsing so it’s proven the market will bear it, and anyways I have a board meeting coming up so rinsing just has to go. Rinsing’s dead now. We’re repeaters. Our company is called repeatli. We’re disruptive like that.
AI — LLMs — are the culmination of this grand, anti-human project of alien-making.3 Previous primitive efforts such as social media had to make do with taking actual human expression and warping it; algorithmically distributing the worst, most hate-making forms of it; interspersing it with ads and sponsored content and — why not? — yet more ads. LLMs allow humans to rid themselves of the expression itself.4 Nothing need be written or drawn or even photographed anymore; just let the AI mimic the beauty of your mind! Isn’t it so efficient? Isn’t it so wonderful to be freed of this burden of creating meaning through the struggle of thought and free expression?
The idea that this all would be fine if only it were less wasteful in terms of resources is completely divorced from any value system to which I could possibly relate. I just cannot accept that one believes this. No level of efficiency and no reduction in resource usage is an appropriate salve for the literal dehumanization of all human expression!
It may seem that I am spitting in the face of would-be allies, but that’s only if you view “disliking AI” as a core, uniting principle in itself. It isn’t! Disliking AI is the consequence of core principles, and here’s the rub: I cannot trust the intentions of people whose “principles” see them slavishly devoting themselves and their engineering expertise to improving the efficiency of one of the most privacy-violating organizations to ever exist. Whatever principles such a person has, they are surely at odds with my own. I’ve no idea what leads them to dislike this newest technological hell, but something tells me it is a fickle belief which will surely evaporate when concerns of theirs which I don’t care about5 are resolved to their satisfaction. And the dissolution of beauty and of togetherness and of struggling through thought itself will carry on; aided and abetted by eager little engineer-beavers; dutifully and self-assuredly gnawing at the trunks of all meaningful relationships; building their well-crafted dams to control the flow of information and attention and the interests of humanity as a whole.
I get it, though: the stock options are a pretty sick deal.
Fellow architects of software; that is: mercenaries trading in human indignity.
Rob Pike is merely a useful hook here; he does not matter. You could replace his name with literally any engineer who’s worked for such detestable companies as Google and who has the simultaneous audacity to complain about AI.
There are legitimate uses of LLMs which are not anti-human and which I care not to explore in this post. This is a polemic, dear reader. Hop on for the ride.
Such despair I am filled with every time I see advertisements for LLMs which burst with excitement over all the ways we can now completely avoid being goddamn human beings amongst human beings.
I actually quite like reading and writing things such as emails, you assholes! Mistake not your anti-human malevolence for a generalizable trait. Some of us still see value in communication; luddites are we, I’m sure.
Deteriorating in class status and losing the ability to be better than.