"Setting aside the moral arguments—"

You mean the power and water.

"Setting aside the power and water, and—"

Don't forget the industrial-scale plagiarism. The brazen theft.

"Setting aside the copyright fuckery, the power and water, and—"

Don't forget the maniacal, suicidal inflation of the bubble. Arguably the greatest single mis-allocation of resources in history, aside from war.

"Setting aside the financial madness, the copyright fuckery, the power and water, and—"

Don't forget the willful destruction of creative livelihoods, the willful destruction of education itself.

"Setting aside the destruction of art, writing, and schools, the financial madness, the copyright fuckery, the power and water, and—"

Don't forget the purposeful degradation of human cognitive capacity. The planned and designed addictive dependency.

"Setting aside the cognitive degradation, the destruction of schools, the financial madness, the copyright fuckery, the power and water, and—"

Don't forget the ghoulish ethical camouflage used to obscure, indeed to erase, the responsibility for decisions in budget austerity, insurance claims, regulatory oversight, medical decisions, court filings, and even real-time combat.

"Setting aside the monstrous mechanisms of official irresponsibility, the cognitive degradation, the schools, the financial madness, the copyright fuckery, the power and water—"

Are you going to say it doesn't work?

"IT DOES NOT FUCKING WORK"

#ai

@jameshowell, and still most I seem to hear from managements and sales is something like:

..... but "everybody" I talked to like it, and it makes you type "things" faster so we'll still mandate use, so we can grow 85% without effort.....and now it can also spend your money/shop without you.

*Sigh*

@jameshowell I can absolutely understand it can do amazing and potentially groundbreaking things in some scientific fields if used right. Things we sometimes aren't capable of.
But I never understood why we need more code. We already have so much code no one probably ever read. And they want even more.
And then I remember they just want more money, and for them more money means more code... But when that money is spent what the hell are we gonna do with so much code no one will ever read?

@chesheer @jameshowell

> I can absolutely understand it can do amazing and potentially groundbreaking things

That's the narrowly focused machine learning though.
Not this Large Language Model nonsense

@benh

You're wrong. It is indeed LLMs that are behind breakthroughs in many scientific areas right now.

Do you change your opinion in the face of new facts, speaking of science?

@chesheer @jameshowell

@SharpCheddarGoblin

I agree. It's indeed funny to see so many read so little.

Easy consumption video here: https://www.youtube.com/watch?v=JvgaZ_myFE4

@benh @chesheer @jameshowell

AI is changing the World Of Theoretical Physics, Fast.

YouTube
@troed @benh @chesheer @jameshowell Maybe I wasn’t clear, I’m laughing at your assertion that LLMs have anything close to the potential to help anything compared to other machine learning. Fuck off sloplord.
@SharpCheddarGoblin @troed @benh @chesheer @jameshowell side note: that is the most conspiracy-theory-the-earth-is-a-hollow-triangle-with-lizard-people-inside-it thumbnail I’ve seen in a ***LONG*** time :P

@benh @chesheer @jameshowell Large language models in the way they are used today are mostly nonsense. There are interesting things that could be done with them, but as someone who actually works with language models for non-chatbot use, I absolutely hate the capture of the term by AI bros.

And I guess at least one person reading the above paragraph will interpret "language models" as something I don't intend, only because of this poisoning of the wells.

@chesheer

A coding LLM is awesome at quickly traversing a large codebase and "figuring out" where the things you want to work on are, and how they fit together. Even if you then don't use it for producing more code.

Most of the vocal anti-AI takes are simply wrong on the facts, but this debate has for a long time now been purely emotional and facts never trump emotions.

@jameshowell

@troed @chesheer @jameshowell

You're so startlingly wrong I don't know where to start, and yes I checked.

Quit playing with matches and give your job to someone competent.

@jameshowell

When the internet is full of AI spam and disinformation, then readers can't trust anything but the big known names.

And employers and publishers can only trust submissions from people they know personally or to whom they are introduced by the old boys' network.

AI is going to be great for gatekeeping the privilege of those already in power & terrible for social mobility.

@FrightenedRat @jameshowell

and this would be different from the past entire existence of humanity how?

@coolcalmcollected @FrightenedRat @jameshowell

um...

took me a minute to come to the answer:

"this is quicker to the result."

@coolcalmcollected

The internet offered a chance to democratise knowledge & boost formerly unknown voices - & thus depart from a history where the media was gate-kept.

AI spam making the internet unusable & many professions unattainable, except for known powerful voices heads off that challenge to entrenched power.

Yes, you could say we are only going back to where we were before, but the way in which AI is being used to accomplish this is worth noting all the same.
@falcennial
@jameshowell

@falcennial @coolcalmcollected @jameshowell

Though I think that in the future some things may be even worse than before - eg because of AI spam submissions, aspiring authors won't even be able to send a submission into a publisher unless they have contacts to vouch for them.

@FrightenedRat

Thiel and his cohorts have always been trying to control how society works because they think they're the smartest ones on the planet. AI is a tool to obfuscate what is seen and heard. Since they own the tools to AI, they present themselves as the grand wizards of reality. (CEO doomerism are just a tactic to bring naysayers into the fold) As soon as politicians climbed on board we were toast. It's the "old boys club" with GPU's.

@falcennial @jameshowell

@falcennial

it's certainly more efficient on their end. why put time and effort into capturing traditional ways of reaching the masses? we all have a tool that can be reconfigured to spew whatever someone else wants us to see and hear.

@FrightenedRat @jameshowell

@jameshowell Yes, yes... But *outside* of all that? My God the *promise* it holds! Just one more model bro 🚀
Stop Gen AI – Mutual Aid and Political Activism

@jameshowell The increasing exasperation matches "What have the Romans ever done for us".

@jameshowell

I had almost this exact conversation with a friend's coworker the other day, we didn't even get to each of these points but I was ready.

This is why it baffles me so much that my coworker's are using this (I'm in education btw). The amount of otherwise measured teachers who can't stop suggesting genAI usage to me. Like any one of these should be a dealbreaker alone, no? Especially for an educator? Like hello? I feel crazy.

@jameshowell Whether it works or not is besides the point.

The usefulness of a product or service does not in itself justify its use. Usefulness is simply a necessary condition for use: if it was useless, you would not use it. Useful means it is possible to use it. It does not mean the use is justified or even justifiable, ethically or otherwise. There are many products that were considered useful but have nevertheless been banned, as their use turned out to be unjustifiable.

@wim_v12e @jameshowell Asbestos, lead and mercury are also very useful
@jameshowell Good one.

I should however add the counterpoint that artists have already sold their souls to large labels, who give anyone but the biggest "artists" - who barely produce anything worth calling so - mere bread crumbs, but lobby to take human rights away from everyone, install rootkits on our computers, and towards a general surveillance state to fight "piracy".

Pirating those, and diluting their influence, is GOOD. Maybe even if done by means of AI, if it needs be (although as you mention the drawbacks are huge too, so this gets difficult). The Music And Film Industry of America (i.e. the MAFIA) must be destroyed.
@divVerent @jameshowell
While I agree with your point, what AI does to media can't be compared to pirating, because the source media that has been stolen is obfuscated sufficiently to be legally 'new' or 'original'
Generally musicians are pro pirating because most of their income comes from tour and merch. The music industry is corrupt and unevenly weighted but AI is not the way to fight it. It's just another burden for artists to bear.
@Shes_Gabrielle @jameshowell My point was rather, does AI harm the large label, or the small artist?

What if we excluded small artists from AI training so only content by major labels is used?

But yeah, it is difficult. My point is just that artists are already being abused by the existing system, AI is not really changing much there. What we need is a breakdown of the existing content industry system. If AI will cause it, so people then can build a new system with hand crafted art, then this would be a good thing. If AI however becomes the replacement, that would be bad.
@jameshowell
Setting aside it doesn't work...

@jameshowell

That's a lot of setting asides for something you think does not work.

Breathe deep!

@jameshowell 🎶
On the first day of AI mess,
The Tech-Bros gave to me:
Co•py•right fuck•ery
@jameshowell i have had multiple people try to get me to "set aside" this stuff and I feel like I'm the insane one?
@amber Let's just say the years since 2016 have not been kind to people, like me, who are sensitive to cognitive dissonance
@jameshowell @amber @steltenpower Isn’t everyone sensitive to cognitive dissonance? I mean, everybody is juggling multiple contradictory perspectives on just about everything nowadays, day in and day out 😬
@jameshowell Think we could get John Cleese to do a dramatic reading?
@ewhac @jameshowell
Could we get a 'what has the roman empire ever done for us' meme-efied version of this, pretty plz??
@jameshowell "ahh no you'll be left behind!!" good. I hope I am. with all the human beings who still have self-respect and brains.
@jameshowell you forgot the bizzare tendency to drive people insane.
@jameshowell oh wait that falls under cognitive degradation nvm.

@jameshowell @FlipperPA I really think it's worse than "AI" doesn't work. Heavier than air flight didn't work, until it did. LLMs *cannot* work. LLMs are categorically incapable of doing almost everything our society is trying to do with them, if reliability is necessary. Bullshit engines should never be trusted with anything, but we are entrusting them with everything.

They're irresistible as accountability sinks in a society that was already avoiding accountability as much as possible, tho.

@jameshowell You forgot the latest - increased heat in the vicinity of data centres.
@annehargreaves It's fun (?) that people keep suggesting crimes that I have forgotten, but it's just that I went past the character limit