David Gerard’s spleen is quite nice.
Why these fucking bozos piss me the fuck off
1 was drafting stuff for this book and it kept turning into short historics where I kept adding “[TK add detail]” and it felt like giving myself homework. That makes for text that bores the reader ‘cos it bores the author.
So no. I’m writing from the spleen here. It’s the only way this can work and have power.
What I hate about AT hype is that it’s by the same shitty bozos who fuck up everything else. They have no approach to the world other than fucking stuff up with money and power via technology.
As a technologist myself (a Unix/Linux system administrator for a few decades), I’m even more pissed off because the technologies are actually interesting, They do things! You could do good things with them! Even the generative stuff, you could play with it and make interesting things!
But no — these bozos being who they are, all they can think of is how to turn it to abuses. Machine learning is for systemic bias. Generative Al is for reducing artists’ labour conditions.
And the power consumption, my God! These bozos were bad enough when they were pushing crypto, and in Al they’ve even managed to replace the ghastly power waste!
Al is not about technology — it’s about power over you.
That all rings true. The technology is interesting and potentially useful, the problem is the techbro cult that is monetizing it all.
Here’s an interesting point. AI used to be marketed as “Expert Systems” back in the 1980s which faded away in the 90s, according to Wikipedia.
In the 1990s and beyond, the term expert system and the idea of a standalone AI system mostly dropped from the IT lexicon. There are two interpretations of this. One is that “expert systems failed”: the IT world moved on because expert systems did not deliver on their over hyped promise.[38][39] The other is the mirror opposite, that expert systems were simply victims of their success: as IT professionals grasped concepts such as rule engines, such tools migrated from being standalone tools for developing special purpose expert systems, to being one of many standard tools.[40] Other researchers suggest that Expert Systems caused inter-company power struggles when the IT organization lost its exclusivity in software modifications to users or Knowledge Engineers.
There are reasons it became less popular as a marketing term.
- Expert systems have superficial knowledge, and a simple task can potentially become computationally expensive.
- Expert systems require knowledge engineers to input the data, data acquisition is very hard.
- The expert system may choose the most inappropriate method for solving a particular problem.
- Problems of ethics in the use of any form of AI are very relevant at present.
- It is a closed world with specific knowledge, in which there is no deep perception of concepts and their interrelationships until an expert provides them.
Sound familiar?
I’m having trouble interpreting the second clause. Maybe replace should be replaced by increase?
…or replicate…
What Does the Spleen Do?
I was just reading an article about the architecture of AI data centers.
For machine learning, they use around a 100 layers of data processor arrays. Each array can be 1,000 processors.
The number of transistors, the basic unit of computing is enormous. It’s in the trillions. Per AI setup.
These AI data centers inherently consume huge amounts of power.
A small one can use 20 megawatts.
A large one can use up to 200 megawatts.
A 100 megawatt power plant can power a small city.
This is brute force computing.
The human brain is an energy hog, and uses a whole 20 watts of power.
“The human brain is very energy efficient, using only 20 watts of power to perform a billion-billion mathematical operations per second. This is equivalent to the amount of energy used by a computer monitor in sleep mode.
Size: The brain is only 2% of the body’s mass, but it uses 20% of the body’s metabolic load.”
You want spleen? Ed Zitron has you covered there today… Never Forgive Them:
[Warning: long!]
One of my time killers is scrolling youtube shorts. It seems like half of them are AI generated. I scroll as soon as I hear a mispronounced word. The other day I saw one about women’s basketball. It pronounced the WNBA as wunba.
We are all in a state of shock and despair over how the election went.
The anti-humans won and are in control.
Things aren’t going well in the USA for the average person. The agenda of the MAGAts is to wreck the USA and make things worse.
I decided I wasn’t going to spend a lot of time trying to understand it.
It’s like trying to read the mind of a malaria protozoan or flatworm liver fluke.
Why bother?
Nevertheless, in passing I’ve made a low effort attempt.
Our standard of living per hour worked has gone down a lot in the last two decades.
The main expense of a family is housing and house prices are at an all time high in terms of affordability.
The second big expense is cars.
Also at an all time high in terms of affordability.
College. Same thing.
Private secondary school Same thing.
Medical care. Same thing.
We also have expenses that didn’t even exist 30 years ago.
Everyone needs a cell phone and internet access.
Job stability is also nonexistent.
Jobs come and go and employees are all fungible and expendable.
Our average lifespans are going down, a sign of a distressed society.
Follow the money.
Life is hard for the average person and getting worse, not better.
People want change, they just don’t realize or care that things can change and get even more dismal, not better.
Some data.
Relatedly I was pondering how anatomy and non-literal language use related to the spleen. I couldn’t think of any emotional language of the spleen. I forgot this reference.
Now is there any emotional language of the pancreas?
larpar: that seems to still be a dead giveaway. It’s funny seeing people argue that nonono that’s all real when someone who has no trouble pronouncing town names in Ukraine (on account of being Ukrainian) pushes a video where all the town names are horribly mangled.
@Dunc – I was just going to recommend Zitron’s podcast “Better Offline”! He’s one of the best tech journalists currently working when it comes to cutting through the hype, and naming and shaming the bastards responsible.
I don’t think he’s had David Gerard on his show yet but I hope he manages it soon. Just hearing those two bounce off each other for a half hour or so sounds immensely cathartic.
What PZ is pointing to is that AI really stands for Actual Idiocy.
In the 1970s we had ‘spurt systems’. Then in the 1980s they devolved into ‘EXspurt systems’ and they all sucked vacuum.
I went to a 2 day conference at UCSC in the 80s about “expert systems” and how they were artificial intelligence that works just like a human brain. The one critic pointed out, among other things, that expert systems weren’t how experts actually solved probs but how they explained how the prob was solved. In other words, it was the justification for the fee charged, not how the work actually got done.
This was all tied up with “5th generation computing” which the Japanese were so far ahead of us on and pouring so much money into (see this slide? just ignore that 3rd column marked “IBM” and you’ll see the Japanese are spending so much more that we are). In less than 10 years (that’s mid-1990s it happened apparently) we’ll have artificial brains that work just like human brains; well, the Japanese will have them and we won’t, because you’re not throwing money at the guys who say that a big bunch of facts tied to a bunch of “if, then” statements is how the human brain works. That’s why Japan now controls all of our computing systems, networks, and products today.
They just move the goalposts when their predictions are shown to be BS.