General Discussion
Related: Editorials & Other Articles, Issue Forums, Alliance Forums, Region ForumsI was just informed from my boss and HR that my entire profession is being automated away.
Last edited Tue Nov 19, 2024, 01:38 PM - Edit history (2)
Not me, the author below.
I was asked to change the URL to this:
https://www.reddit.com/r/ChatGPT/comments/1guhsm4/well_this_is_it_boys_i_was_just_informed_from_my/
Thanks, DU'er native for pointing this out. Safer.
For the last decade Ive worked in local news and have garnered skills I thought I would be able to take with me until my retirement, now at almost 30 years old, all of those job opportunities for me are gone in an instant. The only person thats keeping their job is my manager, who will overlook the system and do maintenance if needed. Thats 20 jobs lost and 0 gained for our station.
We were informed we are going to be the first station to implement this under our company. This means that as of now our entire production staff in our news station is being let go. Once the system is implemented and running smoothly then this system is going to be implemented nationwide (effectively eliminating tens of thousands of jobs.) There are going to be 0 new jobs built off of this AI platform.
There are people I work with in their 50s, single, no college education, no family, and no other place to land a job once this kicks in. I have no idea whats going to happen to them. This is it guys. This is what our future with AI looks like. This isnt creating any new jobs this is knocking out entire industry level jobs without replacing them.
That's the entire post. You may be interested in the 467 replies. Company is not given.
And what about HR? How many does it take to "HR" a handful of people, assuming more than that one manager is kept? One place I worked, HR was outsourced.
And then management?
On edit, I deleted a duplicate post caused by a network hangup.
msongs
(70,170 posts)kevinore
(47 posts)AI is intended to get rid of the cost of human labor making you and the rest of us obsolete. It is not a new problem, we have ignored it. From garbage collection to medical care, AI can replace humans for most, it not all tasks. The question is, what happens when we are all unemployed?
PeaceWave
(941 posts)He or she was suggesting that it was fairly easy to get re-educated in AI skills and re-enter the workforce.
soandso
(1,155 posts)Yes, different occupations will come about but not as many as are lost.
I saw a video with Yavul Harari (I think that's his name, Israeli guy) talking with Klaus Schwab (an actual fascist) and Harari, while discussing AI putting people out of work, said "What will we do with all of these useless people? Video games and drugs, I guess." That is how they view humanity.
One possible upside is that people become more creative and learn trades that AI can't replace. We need people who can do shit, like plumbers, electricians, carpenters. Those trades are aging and not nearly enough new people are learning them.
jimfields33
(18,837 posts)I think some can definitely go into those from some of the AI job Enders. Trades people make great money.
ProfessorGAC
(69,851 posts)A 50 something person is not going to able simply transition into skilled trades like pipe fitting or electrical.
It takes quite some time to get good at those things, and what employer will as a first choice, take a 55 year old apprentice?
Also, as jobs get eliminated, how much new construction of commercial locations will be lost, meaning the demand for that work drops?
Finally, as I presume you believe in supply & demand effects, do you really think they'll still continue to make "great money" when supply exceeds demand?
It seems you didn't think your comment through quite enough.
jimfields33
(18,837 posts)I would at 55 (I am) in a heartbeat if I lost my job. No room for vanity when you need food on the table.
soandso
(1,155 posts)and AI isn't going to climb under your sink and replace broken pipes so these are not only desperately needed occupations but pretty darn safe from being eliminated.
usonian
(13,772 posts)AI will probably plateau as it runs out of training data, currently everything in the planet. And it will start eating itself as the internet fills up with AI-generated data.
As it commoditizes, it will offer no competitive advantage to companies over other companies.
And advanage will be because of people ...
Who are gone.
Tadpole Raisin
(1,499 posts)with buildings/rents and the cost of employees.
Take away those costs and profits soar, at least in their projections.
The costs to us have yet to be realized. Im not one for exaggerating consequences of changes in the work force. Weve seen it many times over the years.
This, taken to its furthest point, makes some of those futuristic dystopian movies not so far fetched.
I would not want to live in that world.
Edited to add: I cant recommend your post enough and I am so sorry this has happened to you!! (Yes I know it wasnt you Usonian).
hatrack
(60,914 posts). . . . .once AI eats our jobs.
Native
(6,550 posts)DSandra
(1,260 posts)Things are looking super bleak for many in America...
calimary
(84,306 posts)What happens when were drowning in unemployment statistics? What happens to all those PEOPLE??? People who have legitimate needs?
Nobody knows, and it seems like even fewer people care.
Delphinus
(12,140 posts)or re-education camps?
Cirsium
(796 posts)"Free market" apologists talk about consumers and workers as though they are two different groups. Of course that is not true.
What happens when we are all unemployed? Then we can't buy anything, and a major depression occurs, because if we can't buy anything they can't sell anything. We will face hunger, homelessness and despair. The fat cats will regroup, buy up everything dime on the dollar and build yet more wealth for themselves and their heirs. Then they will hire us back to produce the things we need to survive, while they skim off most of the value we create with our labor and put it into their own pockets. That will continue until the entire global population has been reduced to penury and the environment has been utterly destroyed.
highplainsdem
(52,322 posts)Rick Beato talked it this in one of Rick's videos several months ago.
soandso
(1,155 posts)and really like him.
A while ago I saw a video about AI music and it was stunning. The music was very good and could easily replace most of the pop and country music on radio.
highplainsdem
(52,322 posts)being sued for it.
See my long reply below.
soandso
(1,155 posts)I know that AI learns by vacuuming up what's out there and then generates something else and I doubt all of the lawsuits will stop it.
FalloutShelter
(12,746 posts)AI art is flooding the market, Amazon has had to make a rule that a single author cannot post more than ten AI novels a day on the platform, so bye bye literature
Not to mention any other job that can be axed.
AI will not have to become Skynet to kill us.
highplainsdem
(52,322 posts)guarantees they're AI slop. And the last I heard they still weren't labeling that stuff AI-generated. Amazon makes money selling AI slop, no matter how bad it is, and they often let fake authors using AI get away with using names similar to real authors' to confuse readers. And there are countless AI-generated unauthorized biographies getting a lot wrong, AI summaries of real books trying to cash in on that book's success, etc.
Very irresponsible of Amazon.
FalloutShelter
(12,746 posts)BUT.... also correct in that they do nothing to label this content.
Delphinus
(12,140 posts)Mossfern
(3,170 posts)does that information need to be disclosed?
Jose Garcia
(2,845 posts)Those musicians lost their jobs after the coming of the talkies.
All of this has happened before and will happen again.
Native
(6,550 posts)Silver Gaia
(4,847 posts)Is there no union for these people? Unions can fight this shit and win, just like they did for writers and actors.
regnaD kciN
(26,591 posts)They can't do a damn thing when they're all being fired at once. AI doesn't belong to any union.
Silver Gaia
(4,847 posts)And of course AI doesn't belong to a union. Yet, anyway. Just wait. If corporations can be people according to SCOTUS... it won't be long til AI are declared people, too.
getagrip_already
(17,430 posts)Right now, AI is "free".
How long do you think that will last?
With the flip of a release, license terms will change, and cost models will be introduced.
AI will become very corporate, and very expensive.
They may even begin to demand derivative rights to their creations, just as an artist might.
This is the phase where the dealer gives you free crack.
TheKentuckian
(26,181 posts)It isn't as much of a "thing" for most workers as some seem to think.
Silver Gaia
(4,847 posts)so far, the only successful fights against AI takeover of human livelihood were the Hollywood unions that fought and won last year against AI taking jobs of writers and actors.
There's a lot more I want to say, and I had typed out a long post on my phone, then accidentally touched the wrong thing on my screen and it went *poof*! Damn.
But I will be back later... I need to vent about what I am seeing first hand that AI is doing to education. It ain't pretty. It's ugly and ominous.
Ursus Rex
(271 posts)... and that's been demonized since they were formed, more and more aggressively the last few years.
Otto_Harper
(702 posts)that with centralized control, and zero humans in the loop, news is even more under the tight control of whomever is pulling the strings from back in the shadows.
wnylib
(24,373 posts)in his prediction of a dystopian world in 1984. Instead, it is 2024.
Steven Maurer
(493 posts)Ownership tries this all the time with new technologies, and it never works out.
There is a lot of suffering while they figure this out though.
PeaceWave
(941 posts)questionseverything
(10,138 posts)soandso
(1,155 posts)I'm talking all of reality. I don't hold that view because a simulation would have to be simulating something else and I don't think the something else is capable of being simulated. It's way too unknown and complex.
slightlv
(4,325 posts)misanthrope
(8,223 posts)all they need do is sit back and wait. We are well on our way to offing our species along with numerous others.
Xavier Breath
(5,003 posts)Doodley
(10,360 posts)hours for the same pay and can enjoy a lot of leisure time. Instead, there will be a lot of pain and suffering under the Trumo regime.
highplainsdem
(52,322 posts)I'd see all too often on Reddit early last year, where fans of AI were dreaming of spending a few days or weeks learning to be "prompt engineers" immediately making salaries well into 6 figures. And where copywriters and others who were already losing their jobs to AI were taunted by AI fans telling them that meant their work was inferior and they deserved to lose their jobs. The AI fans kept up with that delusion even when copywriters explained their old bosses or clients admitted the AI output was inferior, but said they couldn't ignore the cost savings.
On Twitter, I've seen even teachers who once thought AI would help them finally start to realize that the ultimate goal is AI completely replacing them.
And on Bluesky, I just saw John Scalzi (science fiction writer and former SFWA president) post about Harper Collins, a major publisher, having made a deal with an AI company.
See these Bluesky posts
https://bsky.app/profile/scalzi.com/post/3lbaipqnr7c2f
https://bsky.app/profile/scalzi.com/post/3lbaips3i3s2f
and he linked to this, by Drew Broussard
https://lithub.com/harpercollins-is-selling-their-authors-work-to-ai-tech/
Kibblesmith told them no in clear and immediate terms, but here are the screenshots of the offertheyre worth reading for yourself.
To call this bleak is an understatement. And of course there is concern that these AI models may one day make us all obsolete is a truly absurd line to hear out of a publisher engaged in actively feeding one such AI model and Im struck (to put it gently) by the classic union-busting energy of These terms have already been negotiated and agreed to by several hundred authors, so individual negotiation at this point isnt possible. And a $2,500 flat feeper title, so really its $1,250 to Daniel and $1,250 to his illustrator on this bookis an insultingly small amount for *checks notes* the entirety of your intellectual property to be chewed and regurgitated ad nauseam for as long as theres a greedy capitalist there to push the generate button.
-snip-
Generative AI is ALMOST ENTIRELY about stealing humans' work and knowledge so the companies producing the AI models can profit from that theft by selling mimicry by AI of that work and knowledge. There might be a few people and a few uses that are more altruistic. But what's behind the genAI bubble (with countless billions of investment capital that otherwise could have gone to better use, and with companies trying to add AI to every device possible, and to habituate people to AI slop everywhere) is the aim of getting rid of those pesky human creators who want decent compensation and recognition and even respect that their already wealthy overlords don't want to give them.
And every person who goes along with this - who lets the corporations think their AI slop is just fine, or who amuses themselves playing with AI because wow it makes them feel creative and smart so who cares that it's based on theft of other people's intellectual property by a company that doesn't give a damn about anyone - is just hastening this atrocity.
If people are forced to use genAI for work or even school, they may see no choice. But they're being coopted into a capitalist process based on theft and intended to replace them as well. Silicon Valley's new dream, as Sam Altman has said, is a billion dollar company with no employees, just the human founder/owner/CEO assisted by AI.
usonian
(13,772 posts)Last edited Tue Nov 19, 2024, 12:04 PM - Edit history (1)
But it runs out of data and plateaus, AND then everyone offers it, so it's like the 10 cent plastic bag.
https://www.democraticunderground.com/?com=view_post&forum=1002&pid=19739922
And the advantage is in people ...
Who are gone.
highplainsdem
(52,322 posts)AI peddlers like Sam Altman want people to believe they're close to, a superintelligence that might destroy us but will more likely, according to the AI peddlers, be a godlike benefactor of humanity, not only solving all our problems from saving the environment to making us immortal, but bringing us unimaginable abhndance for everyone, with no need to work.
And yeah, a lot of the AI peddlers use drugs and/or seem to be in a cult that's much more religion than science.
And while they try to keep people transfixed by that imaginary AI heaven - with occasional warnings about AI hell and humans being wiped out if we don't have the AI peddlers there to align AI with humanity - they're trying to hide or at least downplay all the harm genAI is already doing. And to get governments to support AI, whether it's allowing them to get away with the greatest theft of intellectual property in history, or subsidizing the building of giant data centers that will use as much electricity as millions of homes.
electric_blue68
(17,977 posts)to a story.
Off, to Google, I go.
ETA
A 1967 poem by Richard Brautigan
"All Watched Over By Machines of Loving Grace"
(Utopian leaning, I think)
highplainsdem
(52,322 posts)Thanks for the reminder. Hadn't read that for quite a while.
I really don't think we can count on loving grace from machines watching over us in a world where Trump, Musk and Putin have a lot of power.
electric_blue68
(17,977 posts)Funny how I thought it was a SF story - alrhough....
an SF author might have quoted the poem's title underneath the title of their own short story. Maybe.
highplainsdem
(52,322 posts)the more common view, or ironic and dystopian.
I did some quick googling to see if I could find any mention of an SF story or novel quoting the poem, but didn't find anything.
Silver Gaia
(4,847 posts)And I have so much to say about this! But it's late and I'm too tired to try to put some coherent thoughts together right now. I hate AI. I see NO real benefit to humanity from it. It will destroy us. IF we allow it.
highplainsdem
(52,322 posts)of data, but generative AI, the type getting almost all the attention now, is much more harmful than beneficial.
I started posting warnings about it not quite two years ago, but the harm it's done since then has exceeded what I expected in this short time.
hatrack
(60,914 posts)So that I'd have time to paint, write, compose, practice, garden.
Nope. AI's destroying my ability to make a living as a painter, a writer, a composer, and the house and dishes are still there and still dirty.
highplainsdem
(52,322 posts)was talking about what jobs will be available for people in the future, and all he could suggest was that some wealthy people might want human rather than robot caregivers.
usonian
(13,772 posts)when I still can't type "and" on an iPhone without it turning into Abba.
THE IMPORTANT STUFF IN LIFE, like those dishes, emergency repairs and getting speaker wires in phase, are left to dumb-shit humans.
signed:
Dumb Shit Human
who grew up reading sines and cosines and tangents! from a book.
mucholderthandirt
(1,147 posts)A major self publishing site, Draft to Digital, sent out a survey about authors on the platform allowing "AI" to use their works for training, with some unspecified but minimal amount of compensation (a few cents per book used, basically, one time fee). The response was overwhelmingly NO!, and the project was tabled. For now.
We're already fighting a zombie horde of people who think "AI" is going to write great books for them, so they can sit back and collect their millions in royalties. It's probably mostly on Amazon at the moment, since they have a system that's been pretty easy to game, with their KU program, but it's happening everywhere.
No one seems to understand that there aren't going to be easy jobs for the millions of people who will be put out of work by "AI". No one seems to care that when you are unemployed, you aren't buying stuff, like food, houses or cars. You aren't going on those great vacations. You aren't buying kitchen appliances, clothes, or paying for medical care.
I see a bleak future ahead, not only due to the incoming fascists, and not just in America. This shit is happening all over. There's no way to stop it. When the population is dead, the rich will have their paradise, run by "AI", all the work done by machines. Where they think they'll profit, I have no idea. Sell endless things to each other, I guess.
Stop this dystopia, I want to get off!
/rant
/first post in a while
highplainsdem
(52,322 posts)has benefited from a lot of them.
When he started out, he would have had to type (or have someone type for him) a clean manuscript (redoing a page if there were too many errors rather than a few that could be corrected with white out or correction tape), probably with carbon paper for his own copy. He would have mailed that out with a cardboard backing so the manila envelope didn't get crumpled, and he would have sent it out with an SASE, a self-addressed stamped envelope, so his manuscript could be sent back to him if (1) it didn't sell, in which case getting it back in good enough shape to submit elsewhere was important to avoid retyping, or (2) the editor could return it after marking changes they wanted.
Editors, if they paid decently at all, did NOT want simultaneous submissions - did not want to waste their time reading and maybe jotting down editorial changes they'd request, only to find out the manuscript had already sold elsewhere. They wanted to see a clean typed manuscript that didn't look like it had already been sent elsewhere.
That need for a clean typed manuscript probably reduced the number of submissions editors got. It required more actual work.
Later editors began to accept photocopies of manuscripts, but they still warned against simultaneous submissions, unless they paid so much less than other markets they were desperate.
Word processors, especially word processing software in PCs, made things sooooo much easier.
But you still needed a decent printer. Editors did not want to ruin their eyes trying to read anything from a dot matrix printer. Inkjet printers were better. Best of all were laser printers, but they weren't cheap. The first laser printer I bought for myself, in the 1980s, cost me $1,200 - equivalent to almost $3,500 today.
And you still had to snail-mail manuscripts. You needed a manuscript.
It's so much easier today, with online submissions.
And ALL that change has been for the better, in terms of giving writers more time for actual writing as opposed to producing manuscripts.
But at least before genAI, actual WRITING was required. Unless someone was stupid/dishonest enough to plagiarize or they had a ghostwriter, you had to come up with the words for the text you were selling.
With genAI, that's no longer necessary. Thanks to AI companies stealing all the writing they could to train their AI, any lazy nitwit with a half-baked idea can give an AI text generator a prompt, and the AI tool will churn something out. What they churn out might contain words the fake writer has never used and might never even have seen before, but some of those nitwits will still try to claim it's THEIR writing because it was their half-baked idea in the prompt, and they'll think they deserve to make money off it, and they'll add to the tsunami of AI slop flooding both editors at traditional publishers and self-publishing platforms.
This is NOT a technological advance - not in any way an improvement - for real writers, editors trying to find good writing, or readers who are going to be offered more and more AI slop.
This is a disaster.
It's also a disaster for teachers trying to teach writing, and for students who - if they use genAI - will miss out on learning both the reasoning and the creativity that learning to write will teach.
But the AI companies don't care. And the non-writers trying to exploit text generators don't care - about either that damage done, or the immeasurably great theft of intellectual property for the initial training of AI.
In some cases AI companies are striking deals with publishers, not because they really want to compensate writers and publishers for the real value of the work that was stolen, but because they hope showing some minor effort to provide some who are being victimized with some trivial compensation will stave off lawsuits and help with PR. And desperate publishers might agree, in the belief that the work will be stolen anyway and they have to strike a deal with a devil who just wants them out of the way and will hasten their demise. And the writers will get hurt most of all.
But hey, there's a legion of fake writers with half-baked ideas and no interest in actual learning and work who will be thrilled with this...until they figure out the AI companies gaining control of publishing don't need fake wannabe writers, either.
At least not if enough consumers are fine with AI slop and don't care if human writers and artists are destroyed.
electric_blue68
(17,977 posts)and AI.
I hung out with a particular sub set of amateur writers who loved writing back in the Mid late '80's - early mid '90s. A few became semi-professional, and and some fewer professional so I feel for writers.
I've written for my own entertainment, and have been graced by a few friends w whom I shared some; who are/were much more versed in writing, and Lit than I was - praising my ability to do really good dialouge. (blush!☺️ )
highplainsdem
(52,322 posts)writers, little different from word programming software or even spell-checkers. All it does, according to them, is make writing easier, and if you don't think writers should use AI, you probably want them to go back to quill pens and make their own ink.
AI fans defending image generators will say just as silly things about artists rejecting those - oh, those Luddites must want artists to make their own paints. I read one article where an AI-using artist compared artists who wouldn't use AI to 19th century doctors who didn't believe in germs and wouldn't wash their hands (crazy comparison, I know, and she was widely ridiculed for it on Twitter several months ago).
They make these stupid comparisons because if they don't try to present genAI as just one more tool they're using, then they have to deal with the reality that THEY aren't doing the writing or creating the art, any more than they would be if they gave a few directions to another person to do it for them.
Real artists and writers know the difference, so their criticisms have to be labeled fear of change.
As if writers and artists haven't already dealt with and welcomed lots of changes. Ones that were for the better.
GenAI isn't.
That's cool that you had a writing circle back in the '80s and '90s, and encouraging friends...whom you should have listened to, and submitted your writing to publishers. Still possible, but now, dammit, those publishers are often overwhelmed trying to find real writers among what mucholderthandirt perfectly described as "a zombie horde of people who think AI is going to write great books for them, so they can sit back and collect their millions in royalties."
If we don't have widespread rejection of genAI soon, so much human talent will be lost. I found out this year that my favorite niece's oldest son, who just turned 16 and has been dreaming of a career as a videographer, was advised by his dad, who works in IT, not to count on many humans still working as videographers in a few years. It hurts to think of kids wondering if any of their dream careers will be left, as genAI takes over.
electric_blue68
(17,977 posts)Sheeesh... what a qualitative difference.
Yeah, olderthandirt's pithy, true comment.
Actually the people i was hanging out with through letters, and later on line were non-profit in that they were playing in other SF Creatives' worlds by writing stories, and making printed zines bc they loved this characters, and worlds basing their stories there but most importantly sold them for just enough money to recoup printing, and mailing costs.No profit.
Some often created their own additional characters, and worlds that either got spun off into to totally their own works without the other's original works (places, characters) in there. A few of those writers became semi-pros, or professionals.
I would get stories from these zine editors, and illustrate them. No pay, of course, except a copy of the zine that had my art in it. So I was on the visual side.
As far as the visual GenAI currently...yeah, it's not good stealing other's visuals. There's influence, but not this kind of stuff.
I'm not a great artist, but I range from good to occasionally excellent in certain ways, especially if I got/get to doing more again.
I spent years in an Art HS, Art College, adult Ed art, my own continuing studies through decades. Started pickingvp digital art along the way, too.
______________________________
As for your ?grand nephew... maybe it'll be a semi-professional thing. I think there'll be some market, but more competitive for videographer, as well as all other creatives.
Hopefully, in some fashion, more will rebel in some ways to stop this type of AI infringement to compleat overtake.
highplainsdem
(52,322 posts)has to be quite good if you were asked to illustrate those zines.
I hope my great-nephew (grand-nephew is also correct) can find something he loves to do. But like you, I'm hoping for a rebellion against AI, for companies (and schools, and individuals) to regret using it, before too much damage is done.
electric_blue68
(17,977 posts)There were fan artists who basically copied headshots.
There was one who was a professional in totally field (teacher, or ?geologist) who was a wonderful artist! Don't know if she ever had any schooling in Art. If not; then a total natural, self taught with a distinct style.
Another one was quite good, but never found out her background.
One who became a friend, and very good may have had at the very least some training, if not more, and did other visual creative work.
I like these fan artists above who included the worlds, and characters both of the played-in worlds, and the original ones the writers came up with; action scenes, quiet scenes, reaction scenes between characters, landscapes, urbanscapes, spacescapes etc. I did my best to treat each story as a professional assignment. Usually with some, to quite complex fore, mid, and background scenery.
Occasional frustrations; mostly issues of time after the work week to do them, but mostly interesting, and usually fun to create! Editors were happy to have me in the artistic mix.
highplainsdem
(52,322 posts)electric_blue68
(17,977 posts)Found 2 fandoms, met some local NYC fans down there. Then I found a zine showed current zines in '87 back at a SF NYC Con. I wrote to a few editors with sample art saying I'd be interested in doing illustrations.
And off I went. Even going to some cons dedicated to zines. Good times!
Still friends w a few people via FB.
Later found and follow/ed stuff on line.
pinkstarburst
(1,546 posts)with Harper Collins offering to pay $2,500 for manuscripts they can feed into their AI for "training" i.e. they will soon be churning out AI garbage under trade labels and fake pen names that are just AI writing the swill and no one will no the difference once the AI gets good enough.
https://www.theverge.com/2024/11/18/24299882/harpercollins-authors-license-books-ai-training
Tadpole Raisin
(1,499 posts)any requirement to prominently post that is what they are using (just like food companies fought the labeling with use of GMO). They dont think we have a right to know and decide for ourselves.
Republicans will love this.
highplainsdem
(52,322 posts)they knew they had a market in companies and people interested in cheating and fraud. Whether it's individuals pretending that they have skills and knowledge they don't have, and that they did work they didn't do, or companies pretending that what they're offering now is as good as what they used to offer and hasn't hurt any workers they laid off or human creatives whose work they stole via AI.
And yes, Republicans and other fraudsters love genAI.
Ocelot II
(120,813 posts)because everything will be AI, so it will have to start eating itself. What will that look like?
highplainsdem
(52,322 posts)argue that won't happen.
In the meantime, it's polluting our information.ecosystem horribly, and even if all use of genAI stopped now and all uploading of genAI results to the internet stopped now, it would take at least years and probably decades to clean out all that AI slop.
That's how bad it is after only a few years of AI slop from text generators (which really took off with ChatGPT in November 2022) and image generators (which really took off months earlier).
highplainsdem
(52,322 posts)electric_blue68
(17,977 posts)"Locked In" was a really interesting book.
I just looked him up, saw a few more of his I'd be interested in.
slightlv
(4,325 posts)where there are at least two segments of the population... those who continue to use technology and those who disdain and have outlawed technology - thus taking us back to the time and efforts prior to the tech. Anything proffered as "will make your life easier" will inevitably be turned against humanity. That's the upshot of these episodes, anyway... and usually, the tech is made out to be smarter, and more benevolent to humankind. And the technology-dismissing society is portrayed as backward, dirty, dull, etc.
I found these episodes of SciFi shows interesting to consider, but never really looked at them as something that could come to pass. Now? I wonder... how long before we find a group of people "taking it back to the basics" and eschewing technology? I don't think it'll take us until the timing of ST: Deep Space Nine to get there. Some things, we seem to be pushing the envelope on when it comes to timing. Just something to think about... regulated capitalism could make all things work for the good of all. Free for all capitalism only works for those who create it - the billionaires. They really do want all the money in the world. Why? After a certain point, what is left to buy? And what intrinsic worth would it have?
hatrack
(60,914 posts)John Brunner
kerry-is-my-prez
(9,197 posts)soandso
(1,155 posts)IIRC, watching the whole thing, this guy enter his own lyrics into the program and AI did the rest. Video starts about mid way and just listen to this music!!!
highplainsdem
(52,322 posts)There's nothing magical about what happens, and no talent or even real interest in music is required, since these companies have stolen all the copyrighted music they could to create AI to compete with the artists whose intellectual property they stole.
It's completely unethical to use these tools. It's also pathetic and dumbs people down, making them less creative and less likely to learn skills they'd need to make music themselves.
See these videos:
soandso
(1,155 posts)I would guess that some AI programs, like what's being discussed in those videos, will make stuff that's nearly direct copies of the ripped off works but some others may not. I'm thinking as the AI learns more, it will get more original with it's own creations. I know that sounds insane because it doesn't have consciousness or imagination but the more it's exposed to, the more it will gain the ability to generate new and different material. Think of all of the music in the world, old and new, different types, cultural influences, sounds that are unique to some traditional stuff, like Mongolian throat singing, the trills and vibrato in Arab and Indian music, classical, Baroque, electronic, etc. Mix it all up and the possibilities for new output are astounding and more than any human could take in in a lifetime. I share your concerns, though, and I think musicians who want to make a living at it are screwed. Not yet, but it's not that far off.
An interesting aside: I had a book on native American art (pottery, weaving fabric and baskets). An interesting thing happened when anthropologists started collecting pottery shards for museums: the living craft nearly died out because people learn by first copying and then creating something original from what they've learned and been exposed to and they copied and were inspired by what the anthropologists took away. Isn't AI learning the same way as the people? It's collecting data and making something new. I have no clue where this goes. It's mind blowing.
mucholderthandirt
(1,147 posts)There are those who say, how is that any different from how "AI" learns? And the difference is, we are humans. We can think, we're self aware, and we can learn in ways a machine can not. It's not the same at all, and we aren't regurgitating entire snippets of different works to make a "new" one. That's illegal, it's called plagiarism. It's rightfully frowned upon in the creative community.
"AI" is destroying the creative arts, and it's only going to get worse, because idiots are always looking for an easy way to make money without having to learn, practice and do the work to make their own art. Stock image sites are full of "AI" generated stuff. All they've done is give some prompts and let the machine make it for them. That isn't creativity, it's cheating.
But, this is the world today. Humanity is on the brink of becoming irrelevant in any way. We don't need elections anymore, the fascists will take care of us. We don't need artists anymore, the machines will do it for us. We don't need jobs anymore, the state will ... Never mind. We can just die and the elites will be happy.
highplainsdem
(52,322 posts)cannibalizes human culture, harms human artists, and discourages humans from learning to create for themselves. It's disastrous for human culture.
The internet is already flooded with.AI slop, ranging from garbage AI art to garbage articles on websites set up in minutes to garbage scientific and medical articles. This crap can be generated by AI in seconds, by people who might have at most an idle interest in what is being generated.
It's ALREADY causing students to switch majors and plans - to decide there's no point in learning anything.
Generative AI is IMO the most harmful non-weapon technology ever.
And it's being used more and more in weapons, where its hallucinations can pretty much guarantee wrong decisions being made and wrong targets hit.
Silver Gaia
(4,847 posts)AI is one of the things that is destroying us, and it is happening so fast. ... I have many more thoughts swirling around, but I haven't slept enough yet to express them well. More later...
soandso
(1,155 posts)While most AI output is shitty, I'm not convinced it will remain that way, though. If it dramatically improves, all the worse for humans. Those with a strong creative impulse will continue to creative because they enjoy the process but making a living at any art form will be much, much harder. That said, it's hardly limited to artistic endeavors because it will effect everything and I don't even know what that means. There have been accusations that the Israelis are using AI in targeting Gaza. I don't know if that's true but I'm sure it will become a very efficient killer, if it's not already, and there are plenty of human psychopaths who make use of it.
I brought up the fact that there are a lot of people who believe we live in a simulation, a belief I don't share. AI could turn out be the simulation that's coming. It reminds me the Gnostics and their beliefs about Archons and a fake reality or imitation of the real deal. I may have to revisit some of that metaphysical stuff to get a better understanding but I do think the AI assault is ultimately a spiritual one. That might sound pretty out there but I don't think it is. We're looking at human consciousness being substituted with something unconscious.
soandso
(1,155 posts)I just asked Bing chat the following and it's answer follows:
Q. Are you able to create truly original content from the huge amount of information your data base. I mean without so much as resembling what you draw from?
A. I can indeed create original content by combining knowledge from different sources and applying creativity. While all the information I provide is based on my training data and the knowledge I've been updated with, I can generate unique combinations, narratives, ideas, and even creative pieces like stories, poems, or artwork.
It's a bit like how humans create new works of art or literaturethey draw on their experiences and knowledge, but the end result is something new and original. If you have a specific type of content in mind, I'd be happy to create something for you! Just let me know what you're interested in. 😊
soandso
(1,155 posts)A little cheezy, perhaps, but here's what it wrote, nearly as instantaneously as my hitting enter:
Q. Okay. Give me a paragraph or two written as a Russian witnessing the beginning of the Bolshevik revolution and the feeling of dread and terror the witness is experiencing as it unfolds.
A. The air was thick with smoke and the distant echoes of gunfire as the streets of Petrograd erupted in chaos. Shadows danced menacingly against the crumbling walls of once grand buildings, now marred by bullet holes and the ravages of war. I clutched my coat tighter, though it did little to ward off the chill that seeped into my bonesnot from the autumn cold, but from the dread that gnawed at my insides. Everywhere, the grim faces of soldiers and revolutionaries, eyes burning with fervor and desperation, painted a scene of unrelenting turmoil.
People huddled in corners, whispering frantically, their voices tinged with fear and uncertainty. Mothers clutched their children, trying to shield them from the violence and madness that seemed to be spreading like a plague. The familiar comforts of daily lifefriendly market vendors, the laughter of children, the simple pleasure of a warm mealhad all but vanished, replaced by an atmosphere heavy with dread and foreboding. It was as if the very soul of the city had been consumed by the storm of revolution, leaving only a hollow, frightened shell in its wake.
***
I only say cheezy because I've read the real stuff but, damn, this was delivered in a second or less and it's pretty good. I've read lots of terrible AI written articles on fly by night websites and this isn't anything like that junk.
highplainsdem
(52,322 posts)impressive, given the billions of words of stolen writing it's been trained on.
The best thing to do with genAI is reject it. We're all on Team Human, or should be. Those peddling genAI are not.
soandso
(1,155 posts)which I brought up earlier about the guy in the videos, discussing the lawsuits, and artists wanting to be paid for feeding it.
highplainsdem
(52,322 posts)and can give you very different answers at different times. Complete waste of time to chat with them.
soandso
(1,155 posts)I'm just sharing how it responds to what we're talking about and that bit of writing it came up with so fast. It can do it because of having the input from human imagination and it illustrates exactly how the creators who feed it will be replaced by it.
I read with great interest the reddit chatgpt stuff when Microsoft was doing it's beta tests on what it called "Sydney", where the thing was hallucinating and claiming to be in love with the journalist chatting with it. It also told him his wife didn't love him. It was like an episode of The Twilight Zone, both amusing and terrifying. You do know that way too many people will converse with this shit as their "friend", right? Though I never saw the film, I know Joaquin Phoenix played a character who had it for a "girlfriend". We already have a generation completely addicted to their phones and living in virtual reality. This could very spell how the human race ends.
soandso
(1,155 posts)is that the guy in the videos is talking about how he and others will be paid and get a cut of the action, rather than opposing training AI all together. That's not a criticism and I realize people want to earn money for their output, it just speaks to the inevitability of the situation. The beast will be fed, like it or not, so may as well get paid to feed it as it replaces you. It reminds me of back in the 90s when tech forced their American employees to train their cheaper labor replacements being brought in from south Asia. If they wanted the severance pay, they did it. It's like, in the end, you WILL be assimilated.
cstanleytech
(26,993 posts)Mossfern
(3,170 posts)I'm not done with my coffee, but I'll try to come up with some examples later ( if I remember)
One thing is plastic packaging.
cstanleytech
(26,993 posts)Mossfern
(3,170 posts)to what the OP is saying - wait - maybe.
I don't think that candle makers lost their jobs with the advent of electricity- it happened over a long period of time. As a matter of fact there is still a need for candle makers in the 21st century.
I don't think that coal miners would balk at being trained and employed in an alternative to the dangerous job of coal mining. Maybe the owners of the mines would be put out, but not the workers themselves.
cstanleytech
(26,993 posts)The coal miners will also adapt as that's the nature of change.
Mossfern
(3,170 posts)AI stifles creativity and threatens objective reality - it can easily be manipulated and abused.
Students are using AI instead of doing the research or writing papers themselves.
In the day of the dinosaur, when I was a student, it's equivalent of tearing a page out of an encyclopedia and handing it in as completing an assignment and the teacher accepting it.
AI has its uses, but needs control - I don't see that happening.
radius777
(3,814 posts)of intellectual property and will consume enormous amounts of energy (that it won't full pay the costs of) - all while producing a largely inferior cookie cutter product compared to what humans can do.
AI can be a benefit in many fields where there is vast amounts of data (science, medicine, law, etc) info that could assist humans in research etc.
But overall it's a mess and represents an existential threat - a world ruled by amoral tech bros is the one we were already heading towards - and AI only turbocharges this.
cstanleytech
(26,993 posts)Can the toddler write a book? Not likely but by learning it might one day go on to do things like solve problems like a way to build a real cold fusion energy source or a drug that cures cancer.
bluesbassman
(19,823 posts)Its a fair enough statement. Humans are designed to grow, learn, and evolve. However ai short circuits that path by essentially taking humans out of the equation. Yes change is generally beneficial and ai is change, but given all of the potential disastrous outcomes of unfettered ai application, what benefit to humanity to see with this particular change?
cstanleytech
(26,993 posts)Even then current AI is still limited in some ways.
hatrack
(60,914 posts)We wouldn't be losing predictable crop weather, or the North Atlantic overturning current.
We wouldn't have an "information" ecology in which the truth is whatever you want it to be. That assumes, of course, that you can still read sentences of more than six words, as attention spans melt on the altar of the phone gods.
The cave's still there. It's just when we head back inside, finding anything nearby to eat or burn for warmth is going to be a chore.
cstanleytech
(26,993 posts)hatrack
(60,914 posts)It's a choice, by people with lots and lots of money who want more money.
And because they already have lots of money, but can't ever have enough of it, they're choosing to underwrite and deploy this technology.
And if it throws tens of thousands or millions out of work and makes this country a more desperate and miserable place? They don't give a shit, as long as they get more money as a result.
Who else "chose" this? Nobody I know or will ever meet, nobody you know or will ever meet.
cstanleytech
(26,993 posts)And yes, most of the wealthy don't give a shit but the people back then all still managed to adapt and I'm pretty sure people like the coal miners and the people in this article that just got fucked over can find a way to adapt in a world with AI.
Tadpole Raisin
(1,499 posts)its another to do this with the quantity of people in jobs AI can replace and the rapidity with which it can be accomplished.
This could (?will) collapse the economy. Rich CEOs are only looking at this by how much more profit they will make in the modern day world of Ebenezer Scrooge and TSF is Scrooge.
In fairness Congress has been warned about this for years and of course has done nothing.
IF, IF there is one good thing about Dems losing, this is likely going to explode on TSFs, made worse by his known plans.
But oh my god the suffering!!
Silver Gaia
(4,847 posts)UGH. I really do not like AI. At all. It's a thief. Of art. Of music. Of writing. Of jobs.
DSandra
(1,260 posts)Just look at the car market. And many others. Companies have focused on serving the top 20% of society that are still doing good while the bottom 80% rots and fends for themselves.
Tadpole Raisin
(1,499 posts)Selling cars so cheaply?
Who will be able to buy them if I dont?
This is far worse! It will crush people everywhere.
Emile
(29,779 posts)thousands and thousands of good high paying union jobs. You would think with all the company savings, the cost of automobiles would be much cheaper. Predatory Capitalism
Amishman
(5,810 posts)I've made a career in IT, mostly on the development and implementation side. Mostly as a project manager, but lately as an analyst (less stressful).
There are a LOT of jobs out there that are at risk of automation. I'm not even talking 'AI' (which really is a misnomer), but from ten year old tech that just hasn't been implemented.
Why haven't they done it yet? Because these projects are difficult and expensive. Very few companies have their internal procedures fully documented, and lots of nuances and exceptions are simply known to the people doing it. It takes a lot of hours by both the implementation team and the team being impacted to document it and account for it. Plus the team doing the job now isn't always the most cooperative. Tech workers who can implement these projects are in demand and very expensive, and offshoring these projects has a very low success rate.
But the cost of implementation only drops as software becomes more capable, and the savings from the workers replaced only rises as their pay does. It will happen eventually.
Johnny2X2X
(21,745 posts)Seems like you do.
In my field, I am about 12 years from retirement from engineering in the aerospace industry. Automation of engineering activities is accelerating, but I think this industry will lag others because of the safety critical aspect of what we do plus the national security aspects. I think I am safe through retirement. But I would be adding more skills if I were in my 20s right now.
People need to look long and hard at their jobs, learn more about AI, and do an honest assessment on their replaceability. People already in tech are used to doing this.
When automation hit the auto industry, people learned new skills and moved into white collar positions. AI could be the opposite move, white collar workers will have to learn different skills to move back into manual labor jobs.
Amishman
(5,810 posts)Rules can be automated, and the more something is done, the more lift there is in automating it.
A significant part of my job is making sense of human focused systems so they can be automated (or at least streamlined), if that ends up being automatable - then we've hit the technological singularity and all jobs are toast.
Blue_Tires
(55,776 posts)Kid Berwyn
(17,967 posts)Its why the Founders put only one business by name in the entire Constitution: the press. The People need the truth in order to base their votes.
A.I. doesnt vote yet, as far as we know. But if and when it does, I bet its owners come out ahead.
Mossfern
(3,170 posts)People (especially greedy people) do no think out the consequences of their actions.
Some might, but they don't give a shit as long as they have theirs and the shareholders make money.
Ugh!
I hate when I right about these things.
area51
(12,140 posts)in addition to universal healthcare. I don't see that happening in the US, though.
John Shaft
(747 posts)more and more things will become illegal in the coming Lawless Times. Ironically, the gaggle of criminals in charge will create felonies out of thin air for the sake of punishment and free labor.
LucasD
(84 posts)She told me that her company has just started experimenting with AI, hoping to replace her position.
She couldn't go into a lot of detail, but I have to assume that damage metrics would be loaded into
a computer, and historical replacement data would be used to let the AI determine the replacement costs
and value of property. Maybe AI would be used to analyze damage photos?
I can't believe that AI will ever be as accurate as a human in assessing the actual, and especially the hidden, damage to property.
Also, while many people get defensive in her presence due to fear of being low-balled on replacement value,
there are other times when she becomes a sympathetic ear, offering offer words of encouragement during a really trying time.
No AI is going to do that.
Greed sucks.
hunter
(38,919 posts)Her last radio job was at a station very much like the one depicted in the television show WKRP in Cincinnati.
That world is long gone, largely replaced by automation and syndicated content.
XanaDUer2
(13,829 posts)Liberal In Texas
(14,489 posts)This is horrible. The quality of your local news is going to look like crap. The station owners won't care as long as they can sell ads. This is going to amp up the further decline in viewership of your local news. It's going to look like amateur hour.
John Shaft
(747 posts)Midmarket and smaller? Those markets disappeared like the greyhound bus stations and payphones did.
Oneironaut
(5,768 posts)It hallucinates constantly. Companies using it to replace entire departments are going to find out its going to create more problems than it solves.
Prairie_Seagull
(3,758 posts)There are plenty of jobs at walmart. Trade a 35 dollar an hour, many times UNION job in paste up, for an at best 15 dollars per hour as a greeter. Boy oh boy what a bright future.
Don't think this needs a sarcasm thingy but here is one anyway.
jmowreader
(51,438 posts)In the 1980s companies like Aldus, Quark and Ventura created programs that allow you to build a complete print-ready document on a computer, and the old "run out galleys off your phototypesetter and paste them to a layout board" publication model died. And now that everyone outputs plates rather than outputting film and stripping it into flats to burn on a plate frame you can't even use it as a backup workflow.
My sister's first job was doing that. She ran Compugraphic phototypesetters, and she had two machines - one for headlines, one for galley text. Strangely enough, the newspaper she was doing it for still owns both machines. It would cost more to have them scrapped than to just let them sit in the back room. The thing she hated most was setting legals: the galley machine doesn't have a numeric keypad so the thousands of numbers in a legal had to be entered from the top row of the keyboard.
Prairie_Seagull
(3,758 posts)Admittedly it has been a long time. Thanks jmowreader for the correction and excellent description. My wife, years ago did a similar job in what then was known as pre-production
jmowreader
(51,438 posts)I've always worked in the electronic workflow. When I started we had an Agfa SelectSet 5000 imagesetter (Agfa bought Compugraphic decades ago, and the film cassettes we owned for this machine had "Compugraphic" logos) that we could run either single sheets or two-page spreads on. When we bought a six-color press we also bought a Scitex Dolev 4press imagesetter that would run four-page impositions...for our 28" presses one sheet of film covered a whole plate, on the 40" presses we stripped two sheets into one flat.
The best part about Scitex is you could change the screen angles and dot shape for each color right from the RIP, and of course we did. Turns out offsetting the black and yellow dots from their standard angles by about a degree and a half gives you a way better looking print. They also have a KILLER vacuum system on this thing; it's internal-drum and sheet-to-sheet consistency on it is unreal.
The worst part is Scitex equipment is EXTREMELY fussy about humidity - which you wouldn't expect from equipment built in the desert. For some reason all Israeli-made graphic arts equipment is nit-picky about humidity. I've got an Indigo press on my floor right now, and if the RH drops below 50 percent it quits working.
electric_blue68
(17,977 posts)a bendable photo lithography plate now created as a whole vs stripping the films into flats onto a plate frame?
In the early '70s took a photo lithography course at a medium to higher end printing shop from my Art College.
jmowreader
(51,438 posts)Most of them are sensitive to violet light from a violet laser diode. They're fed through a machine that I tell visitors to the shop is "kinda like a laser printer except it's bigger than your car and more expensive than your house." From there they go into a plate processor that makes them usable. Our press requires the long edges be bent; if you have presses with pin registration systems the platesetter can punch the plates for you.
This technology can do about 95 percent of everything I can do on a stripping table and about 80 percent of everything someone who was trained to strip can do...I learned it in desperation because I walked in the shop one morning, got told "the stripper's going to be out for a little while to care for his terminally ill wife and I'm not hiring anyone else; you keep us running." The stripper was out for five months.
electric_blue68
(17,977 posts)electric_blue68
(17,977 posts)Brings back some Art College memories; some pretty vivid, others a bit hazy.
Among the various things I played with was Kodalith film. Set up stuff horizontally in a ?vacuum frame, then tilted vertically. Then I turned away from light source before flipping a switch after adjusting the carbon arc light rods to expose the film to get clear type on a solid black background.
Sounds like an "ancestor" to your violet laser diode!
On a rare occasion I'd be stripping this type into some layout I designed for a Graphic Design assignment, along with that very cool transparent ruby red tape. Maybe an equivalent liquid as well for certain applications when the tape wouldn't work. Memories!
Also making totally high contrast images for usually photo silk screening. Occasionally then photo printing high contrast images.
More memories! 👍
jmowreader
(51,438 posts)The big thing when I was still in a film workflow was Rapid Access film. This could be processed in the same chemistry as imagesetter film.
electric_blue68
(17,977 posts)jmowreader
(51,438 posts)Once theyre processed they work the same but if you have an imagesetter you will want to run your camera film through the same processor, and Rapid Access film lets you do that.
electric_blue68
(17,977 posts)electric_blue68
(17,977 posts)occasions [pu & m] till mid-late '80s the old fashion way.
Mechanicals were usually the higher skilled craft of pu & m for those who wouldn't know.
Although welding a very sharp #11 exacto blade & handle, to cut out the quote marks, apostrophes, and periods (question marks being easier 👍 ) was quite a fine hand/eye coordination skill vs easier moving a single letter, a word, or a strip of a few words as directed by (usually) an editor. 😄
One place I worked free lance 2 wks/month for ?1 1/2 yrs was 🩷 Schoolastic Magazines. Computer graphics were just beginning to filter down.
LAS14
(14,678 posts)Soon the educated and the un-educated will face the same threat of job loss. Maybe we'll all figure out a way to band together to influence the character of this bright new society.
FSogol
(46,502 posts)LAS14
(14,678 posts)What do we know about its veracity?
Edit: I was able to open the whole thread, and there seem to be enough comments with similar stories, and none questioning it, that it's probably true.
usonian
(13,772 posts)Who checks? Nobody.
Money matters more than anything else. It's a societal problem.
Delphinus
(12,140 posts)says it so well.
Chalco
(1,357 posts)My daughter's in the animation industry and has seen hundreds lose their jobs to AI. There's a movement to limit the influence of AI but who knows what's going to happen. Pretty soon there will be no jobs and we will all be paid to live or die.
Native
(6,550 posts)LAS14
(14,678 posts)Native
(6,550 posts)if anyone clicked on it before it was changed, it could be a problem. the original url was to old.reddit. - that's not reddit. when you clicked on it, it brought up a fake site that mirrors the actual Reddit site, so. it still looks legit, but when you forward the link from that site, that's when things start going crazy. Who knows what else it would do - I didn't stick around to find out. If anyone has sent that link to another person, they need to contact them and tell them not to click on it.
Native
(6,550 posts)I clicked on it and the site seems to mirror reddit, but if you forward the link it changes it to redtit. yes, red tit. and if the person clicks on it the flashing VIRUS ON YOUR PHONE CRAP POPS UP.
LAS14
(14,678 posts).... but it contained a very long thread of what looked like legitimate comments. It is possible, I think, to use the old Reddit format by choice. When I clicked again, it took me to the same thread, but in new Reddit format. I'm on a computer. I get not warning from Malwarebytes.
Native
(6,550 posts)I didn't notice it was old.reddit. I let my guard down when I'm on du. Not going to do that again. We should all check the URLs we're clicking on regardless of where they're coming from.
LAS14
(14,678 posts)... as if I went to Reddit.
Native
(6,550 posts)Luciferous
(6,261 posts)give us a bot to "help" us. The thing was so bad that our productivity dropped significantly because we had to keep fixing the bot's mistakes. Now we don't even use it, and they spent a LOT of time and money on it. Automation is not always more efficient, they might learn it the hard way.
highplainsdem
(52,322 posts)LAS14
(14,678 posts)... have to train ourselves to suss out bad sources, even if the content is true.
StarryNite
(10,821 posts)work in the farm fields, construction, and restaurants. Because those "immigrants" stole all those jobs from white Americans.
jmowreader
(51,438 posts)Ndp5
(66 posts)Boston University suspended its Humanities and Social Sciences Ph.D. admissions on the heels of launching a multiyear Artificial Intelligence hiring initiative.
https://www.insidehighered.com/news/admissions/graduate/2024/11/19/bu-suspends-admissions-humanities-other-phd-programs
https://www.bu.edu/cds-faculty/culture-community/join-us/faculty-positions/ai-cluster-hiring-initiative/
The only people Ive met who are impressed by what AI can do are people of stunted intelligence with limited skills. Everyone else is justifiably skeptical and worried AI will give managers an excuse to kill jobs that humans now do capably and bots will just fuck up.
usonian
(13,772 posts)That's a very hard wall to hit at high speed.
I very much look forward to buying used GPU's for a buck or two at the flea market.
Except that I haven't figured out a worthwhile use for them.
But I'm working on it.
Not hopeful that the "Ultimate AI computer" will figure out that it is causing more harm than good by learning from the past data. The past data includes many colossal failures, some very recent.
LAS14
(14,678 posts)Nimble_Idea
(2,421 posts)smoke em if ya got em
o7___o7
(4 posts)I submit to you that the entire AI grift is running on para-religious mania and fumes. Ed Zitron lays out the case perfectly here:
https://www.wheresyoured.at/subprimeai/
In summary, OpenAI et al believe that LLM (that is, large language model) performance scales linearly with training data and processing power,. Therefore, all that they need to do is continue building data centers and scraping all possible text sources to improve their AI models to a "god-like" capacity, after which any bills coming due (such as loans or violent social upheaval or climate change) will be rendered moot by the apotheosis of our new robot overlords.
This is dumb:
1) Turns out LLMs don't scale like that, and even if they did, the industry has burned through almost all of the human-generated texts in the world.
2) As an example of the hole that is being dug by the industry, OpenAI, by far the dominant firm in "AI", is spending money 5x faster than it can make it, and is making up the deficit with fund raising.
3) OpenAI raised 5 billion this year, giving it a nine-figure market cap.
4) OpenAI made 1 billion this year.
5) OpenAI loses money on every API request.
6) Large institutional investors such as Goldman Sachs are openly skeptical.
7) OpenAI has released several not-quite-GPT5's with huge fanfare that have all failed to improve over their predecessors.
8) OpenAI has projected that it must spend trillions (With a t) of dollars to build the physical infrastructure to power the next generation of AI models.
9) But LLMs just don't work. No matter how much money is poured into hardware and training, these LLM architectures will never be capable of more than being "spicy autocorrect."
In the next few months, barring a huge infusion of capital from a nation-state level angel investor, these companies will start falling like dominoes. The only question is how much human carnage credulous idiots like this guy's boss will cause in the meantime by preparing for a magic robot future that will never come, because people need to eat in the meantime.
For a more rigorous, edifying, deeper dive, I highly recommend "On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?" by Drs. Bender, Gebru, Mitchell, et al.
https://dl.acm.org/doi/10.1145/3442188.3445922