I want to remind you: these bastards have spent $1.5 trillion of society's money on this technology. That it's produced *something* should not be surprising. That's 'solve aging', 'cure cancer', 'eliminate all poverty', 'shift climate change into reverse', or 'send a probe to another solar system' money. It's not 'write shit software 20% faster' money.
@jsbarretto actually i really only think it's going to be faster if you're running it without checks. if you're not phoning it in, i found it slower.
I have found that if you are mediocre coder (like me) artificial stupidity toos are faster but at the price of me not learning anything (and thus not getting better) and even worse not understanding my own code... which makes fixing things down the road even harder. It's not worth the trade-offs.
@jsbarretto research I've seen says 10% while those using it become less capable due to cognitive decline, and vastly over estimate the benefits they receive.
It truly is the ultra wealthy creating a new #LLM asbestos that they fully intend to make everyone swallow regardless of the detrimental impact because it gives them more power and control.
It's not about profits, they have plenty of money. This is about controlling a population when people become desperate.
@jsbarretto plus the environmental impacts of the power usage and water usage!
Let alone the other impacts like the rising cost of electronics parts meaning smaller businesses can't afford to start or ship product.
AI is a trainwreck, this is going to be worse than the dotcom bubble when it goes bad...
@jsbarretto 20% shittier, not faster. That definitely changes the equation!
@jsbarretto Solving aging or curing cancer doesn't help the companies funding this tech. LLMs are designed to give even more power to the 1%, it's the perfect scheme, you pay for a tech that makes you stupid and more dependent on said tech. If LLMs could really empower the regular person, they would NOT be pushing it down everyone's throats.
@jsbarretto
what is the antecedent to "this"
a very common error on social media, a pronoun without a clear antecedent
@failedLyndonLaRouchite Swimming pools
@jsbarretto
> these bastards have spent $1.5 trillion of society's money on this technology
@failedLyndonLaRouchite
> what is the antecedent to "this"
Unless you've been hiding deep in the forest for a decade or more (is 'nam over now, man?), the answer to that is painfully self-evident. Do you know of any other technology that bastards have spent $1.5 trillion of society's money on recently (or *ever*)?
@strypey @jsbarretto I have the laRouchie blocked. Not regretting it.
@jsbarretto Well more than that, actually. I was reading how they're using shell companies to hide at least another 1.5 trillion. And I'm sure they're playing plenty more funny money shell games to hide even more debt.
Aug 14, 2026, 3:36 PM@jsbarretto maybe we should cut them some slack it's only "end world hunger" 250 times money /s
@jsbarretto
I've said it elsewhere, and I'll now say it here:
Generative Ai is the Cyclospora of enshittification.
It takes a toxic process to an alarming level.
#Ai #Cyclospora #Enshittification
Aug 14, 2026, 3:48 PM@jsbarretto from their perspective, it's "make all creative work dependent on these six companies" money
@jsbarretto
It's not even 20% faster when you add debug/test/maintenance/changes. It costs more in money & time.
Time to write code was never the problem. It was analysis, design, test, debug, then changes required after release.
@jsbarretto while I totally support your sentiment that this money should have used better, I have the bad news that we’ll probably not be able to reverse climate change with all the money in the world. But we would be able to mitigate it a bit and make the lives of everyone so much better in the progress!
@acid Agreed. My point is that a gargantuan effort has gone into changing what seems, ultimately, to be a fairly underwhelming technology.
@jsbarretto "solving aging" would be a disaster. The younger generations are our only hope.
@jsbarretto literal thousands of cities for displaced people around the world, or computer that leaks poison and can't do math
@jsbarretto In quaint times of yore, people would convert new technologies into products and accept their success based on timing, marketing, and identifying an actual utility.
These big model guys want us to develop the product and rent us the engine, in dreams of big bucks with no risk. Now, mind you, they’re spending investor money, so it’s a race with the runway using circular financing and artificial valuations.
@jsbarretto It's done a lot more like:
- Teach people to think in terms of corporate incentives in the sale and manufacturing of products.
- unmask hyper-local corruption
- Show what Capital has always been capable of, but hasn't done for us.
- a video of Will Smith eating spaghetti that will haunt us to our grave.
I don't know about eliminat all poverty or shift climate change
but in regard to sole aging or cure cancer:
the world currently spends about US $250 - 300 billion a year on pharmaceutical RnD
so no, 1.5Trillion is probably not enough to "solve aging" or "cure cancer"
and those problems aren't entirely solvable by money - they are a bit like putting together a world class sports team, it takes years to get together the right team and train it, but even more so in biomedical research, you need a decade to train someone, another decade to set up a world class lab
Aug 16, 2026, 2:58 PMI don’t think that’s entirely fair.
If you include the off-the-books loans, the number is closer to $3 T.
@david_chisnall @jsbarretto and then there are the 'guarantees' which aren't booked as debt
Aug 17, 2026, 11:16 AMThe obscene scale of AI investment is a perfectly valid criticism of capitalist priorities.
But “imagine what else we could have done with that money” is not a programme. It does not tell us who should be taxed, what should be publicly funded, which AI uses must be banned, how workers and creators are protected, or how the infrastructure is controlled.
“Never use AI” turns structural criticism into an individual purity test. Convenient for shouting at random users; remarkably useless against hyperscalers, landlords of compute, venture capital and governments handing them subsidies.
We need regulation, labour rights, consent and compensation for training data, antitrust action, environmental limits, and hard bans on surveillance and automated life-changing decisions. Not another moral panic pretending that a sticker is industrial policy.