Thread 106197372

>It's literally the Manhattan Project 2.0
>turns out it's a giant nothingburger
Now that the dust has settled, what went wrong? Did they lobotomize for safety it too much?

>Manhattan Project
impotent boomers don't leave the bed without having their wild sexual fantasies tingled
that's all it is

>>106197372 (OP)

>a program that just tells you what it pulls off google

>>106197427
So it's literally BoomerAI

>>106197444
It's definitely such a thing because the doomsday bullshit appeals to boomers. Boomers also love it when number goes up, because it means they've siphoned more resources from the young.

>>106197372 (OP)
grift, always was

>>106197372 (OP)
The goal of GPT-5 as it has been released to us was to enable the elimination of all the other models which had proven too costly, so that this one could give most users very cheap responses with no way on their part to decidedly circumvent that. It's all desperate cost-cutting measures gone about in a totally dishonest way. It seemed it didn't occur to them that depriving people of something they were satisfied with before while replacing it with something noticeably worse which you nevertheless insisted was better, would be a sure way of turning those consumers off and away to the competitors.

So now you've gone from being synonymous with AI to being the scummy one in a field of okay options.

>>106197372 (OP)
More like Theranos 2.0.

>sputnik moment
>manhattan project
What else?

>just 100b more parameters bro
>we're getting AGI right now bro just 100000b more parameters bro

>>106197372 (OP)
They were just wrong about neural scaling being pure magic. And they probably realized this eventually but it was such a simple narrative to tell to investors that Altman wanted to keep doing it.
This recent paper explains it all.
https://arxiv.org/pdf/2507.19703

>believe someone who looks like zuckerberg
oh no i got scammed
>believe someone who looks like sam altman
oh no i got scammed again

>>106197372 (OP)
?What went wrong with our giant slop soup? Did we lobotomize for safety too much? It was gonna be groundbreaking, right?
Fucking kek. Why do normgroids do every kind of speculation, except for speculating that maybe the jews involved were lying from the outset?

>>106197372 (OP)
>What went wrong with our giant slop soup? Did we lobotomize for safety too much? It was gonna be groundbreaking, right?
Fucking kek. Why do normgroids do every kind of speculation, except for speculating that maybe the jews involved were lying from the outset?

Seems to be exactly the same, still spews shitcode you have to manually fix, still helpful though

>>106197372 (OP)
>>turns out it's a giant nothingburger
No, this is actually a worse product...that's significantly retarded.

>>106197372 (OP)
They thought free tier users will not notice the downgrade i they just replaced the existing models with the new faster but dumber models. They do have smarter models they just dont want to give it all away till they are able to scale up to better performing ones. Their IMO Gold winning model is probably the smartest model they have left and now tey have to figure a way out to gett a smarter model than that somehow

>>106198750
>They thought free tier users will not notice the downgrade i they just replaced the existing models with the new faster but dumber models.
And they were wrong. This fiasco really proved that nobody on the board cares about anything other than STEM, and even then were so blinded by money that they were willing to force STEMfags to pay in order to sidestep this garbage.

who scams the scammers

>>106198798
Scammers scam themselves with their own horseshit. You'd be surprised how often they self-delude and start believing their own lies.

Why is /g/ against AI suddenly? just being contrarian?
This place was AI cult central not long ago.

>>106197372 (OP)
Can't forget
>It's so advanced it scares me, I should keep it locked up, but I have to release it because the market demands it

>>106198465
>Just 100 more experts bro
>We just need 1 quadrilion synthetic training tokens bro
>All that is missing is the exaflop datacenter powered by the anti-matter power core bro

>>106199139
AI threads on every board are either the same jeet shills or troon coomers.

>>106199139
the shine wore off, people see it for what it is. now everyone is just waiting for the bubble to burst.

>>106198408
Good ol' Forbes curse.

>>106199139
The slopification of the internet and increased bot activity. Time to focus on quantum computing and drop the AI lolcow

Only Google pro can beat Grok 4 atm, but thats excluding Grok 4-Pro/Heavy version from benchmark.

>>106199564
>people see it for what it is.
The future of coom and a tool that replaces all coders?

>>106200153
And other idiotic fantasies the glavset needs you to believe.

The tech has plataeued. There isn't enough good data left and scaling has gotten prohibitively expensive. All they can really do now is minor tweaks and focuses.

>>106197372 (OP)
at gpt 6 release he gonna be like
>I have completely ascended into shamballa and I have touched the gods, for only $199/mo you can too

>>106198798
nvidia

I hardly know anything, but the way they're trying to reach AGI seems confused. It seems like in an effort to get natural language to work perfectly, they've assumed that the same method could work for knowledge and thinking. There has to be a different analytical approach to the latter. It's the algorithm/method/whateveryounerdscallit instead of the amount of data. What human has ever read as much as GPT? Yet a human 1000 years ago with access to a few books is smarter than GPT in terms of reasoning ability.
They should mimick the human brain. and I don't mean just create a neural net in a creatively layered way. There's some part of the human algorithm or structure missing in these programs.

>>106199583
The face of current america: con artists

>>106200423
But the entire American economy is betting on exponential AI growth and profitability, if that fails to materialize America will go bankrupt and collapse

>it scared me!

fucking faggot
never forget this board shilled that trash for free

>pencil eyebrows
>nose job
>lip fillers
In a sane world, this guy would be in a mental institution.

>>106200643
>never forget this board shilled that trash for free
Because 4chan is swamped with bots and pajeet marketer shills.

>>106198798
Is it really us getting scammed if they are losing money on free AND paying users?
I’m pretty sure I’ve personally made them burn thousands of dollars and I haven’t paid them a cent.

>>106200922
The expensive part is training, you're just feeding the prompt history

>>106197372 (OP)

oh it is a boat then see if z80 can refine it is just roman roman rowboat for 105 men

>>106198476
I mean, this paper is retarded. Two "scientists" predict the future based on some shitty assumptions and even shittier math. Fucking astrology tier. They should quit their college jobs and take up palm reading. Not science.

I do remain hopeful the bubble bursts, if only because a lot of the technology intended for these "child safety" bills leverage AI...

>>106200614
why don’t millionaire billionaire trillionaire gorillonaires just invest in actual quality of life

>>106197372 (OP)
it's so dumb, it sounds so retarded and inhuman now that no student will be able to pass GPT's stuff as his own writing.

I rejoice. Once normies notice this is going to destroy the company

>>106197372 (OP)
there are just hard limits to what you can achieve with the available flops, electricity, data and algorithms we have..

>>106199982
they are all benchmaxxed tho.

we need deepseek to come back

>>106201296
I think in time we'll see there not being much difference between the commercial models and open source as we reach the limits of the tech.

>>106201244
there's more printed money then there are good and profitable companies to invest in.

>>106200614
>>106201354
In the mid-2010's I realize the economy was just a made up circus act, so nothing will ever happen

>>106197372 (OP)
he looks like that cuckold guy from black mirror lol

>>106201296
I haven't been paying attention for a bit, what happened to deepseek?

>>106197372 (OP)
>What went wrong?
Nothing? This thread is just another proof that his marketing is working.

>>106200569
But did you see how much GDP they add to the economy?

>>106203641
>>106200569
do yurpoors normally cope this much here? I know you're poor, but come on, relax and enjoy foraging for food!

>Hello, I am CEO of "company".
>I'll have you know our product is ground breaking and you should invest all of your money into it.
Well, of course, I'll immediately do so!

Why oh why?

>>106203655
why are europoors living so rent free in your head?

>>106203670
I don't think about you at all

>>106203684
why do you have so many images with EU if you don't think about us at all?

>>106203684
Altmans stole a part of your dick along with your future, and you think about yurpoors. Really a cattle

>>106197372 (OP)
>Now that the dust has settled, what went wrong?
A lot of us have been screaming that there's a hardware bottleneck, but our criticism was handwaved because it would presumably be solved by "muh efficiency improvements".

>>106197372 (OP)
>We just need to scale up 100x and it'll be agi bro
>trump-sama please build us a private nuclear reactor to scale
>agi 2026

>>106203713
lmao. rent free.

Where's the IQ cutoff regarding Ai?
No actually smart person I know gives a shit about current AI.
I know quite a bunch of smart people, them usually being above about 130 IQ wise.
Even on that level, AI is not useful for much. If you ask it to simplify a lot of information into a bunch of simple points... You still have to double check it because current AI sucks.
Basically, if you are pro this, you are doing a very simple job.
Also, you are pro something that could overtake your job.
There's no reason to any of his.

>>106203844
you are a massive retard
t. american

>>106203854
To double down, you guys don't even get payed for this.

Where have we heard this before?

>>106197372 (OP)
Here's my conspiracy theory: someone already has access to an AI beyond even your wildest dreams. But they don't use it for scientific research, coding, or writing spam emails.
They use it for information control, to shape the narrative. And they're good enough at it that you don't even notice.
The rest of AI development is nothing but a distraction, to make you keep thinking that AI is still dumb enough that you'd notice instantly if your entire worldview was controlled by a computer.

>>106203844
Dude, if you want free healthcare and a cute European girlfriend, just ask. Come. I'll let you live in my spare room.

>>106198169
>which had proven too costly
That doesn't make sense.
We are talking about a company that isn't profitable in the first place and that gets hundreds of billions of funding.
I don't think hosting costs are an issue.

i think he's just kind of dumb. you could argue he is just a hype man getting eyeballs on his product but he kinda blew his load on this one. nobody is going to take him seriously if he claims anything like this again.

that or he knows the limits have been hit so he milked this one

>>106197372 (OP)
sandwichmen and marketing people should be locked up

>>106203655
All those high paying american positions are taken by slavs working remotely from their basements.

>>106203917
almost everyone on the planet is controlled by an "AI" algorithm already. Every single major web tech company is using deep learning to maximize engagement on their platforms. you train it to train you to spend as much time using the service. think, youtube, twitter, tiktok, amazon, etc. Now LLMs are doing the same thing, maximizing time spent using the product. Why? Ads and information control.

>>106197372 (OP)
Altman is the peak scammer of our life time

>>106204245
I haven't seen any ads in decades.
Neither have people in my inner circle.
In the meantime you say stupid things.

>>106203655
no human needs this much money

>>106204299
Nobody does indeed.
But you also have to include mental illness.
Some "humans" may turn out to be just, actually inhuman, monsters.

>>106204332
A lot of them also only eat kosher food for some curious reason.

>>106203740
The AGI cult labels naysayers like this as "negative thought persons" and shadowbans and blacklists them

Demis wins

>>106197372 (OP)
tHEY NEED AN EXCUSE TO BUILD MINI REACTORS THAT WILL BE THE BASIS OF FUTURE ENCLOSED 5 MINUTE CITIES.

tHEY NEED AGE VERIFICATION TO TARGET THE DEVELOPMENT OF UPCOMING GENERATIONS WITH FINER GRANULARITY.

Sorry caps lock. Not retyping.
:^)

>>106203655
>Higher by $400
Maybe if you include after taxes, but no way in hell a Euro city comes anywhere close to San Fran. Unless you're making 100k a year in San Fran you're a pauper. And over 100k is just lower middle class.

If you went back to 2011 and told me that we were going to have self-driving cars in 2025 I wouldn’t believe you. Even the electric car would be hard to believe because of the greedy oil barons.

>>106204579
Waymo and Telsa both existed back then, and had demos.

>>106204579
>If you went back to 2011 and told me that we were going to have self-driving cars in 2025 I wouldn’t believe you.
We don't.

>>106204280
hence the key first word "almost"
most people still don't use adblockers, bad argument from you anon.

>>106204637
The Tesla Model S release in 2012 and I’m actually surprised that the oil barons didn’t suppress it. Elon was just too rich and powerful this time.

>>106204659
Skill issue, git gud at prompting >>106204659

>>106204703
I guess it's just sad.
Just more reason for me to educate people about adblockers.

>>106204797
Yea, just install it for as many people as you can. The bigger issue is undisclosed, unblockable ads, put inside the content by content creators paid to do it. Often as subliminal as possible. Not just blatant instagram influencers but virtually every account on social media with a following will have people spamming their DMs to say something favorable about x product/service in exchange for money

>>106204834
It worked out fine for me so far, I hardly ever have to watch ads.
I always go for Browsers outside of what most people do, I use Edge, Brave, stuff like that. Combine with ublock origin, stuff like that.

>>106199139
>This place was AI cult central not long ago.
you have to be retarded not to realize that it was all bots, paid shills and a few low IQ technologically illiterate subhumans who eventually went back to /tv/ or wherever the fuck once the hype died off.

>>106197372 (OP)
Rule number 1 of IT sales. Always overpromise and underdeliver.

>>106204055
I work roughly 30-40 hours a week at the very most, 3 days at the office+2 from home (that's "officially"; in reality, it is more like 2d/week from the office at most) at a FAANG. $200k base+around $150k more annually in stock vests (heavy majority of that chunk)+bonus.

The "muh slaves working ungodly hours from the basement" is an overplayed meme. I know some SWEs in the US who work ungodly hours and sacrificing their work-life balance (esp. in quant space), but they are typically doing it for way higher pay and intentionally.

>>106204948
That sounds like you're a retard, working way too much.
Wtf is wrong with you?

>>106205017
~6 hours a day is too much? Mind you, that's not straight work the entire time, that's including lunches and gym and other stuff.

I guess it might be too much for some people, but idk what else to say other than "git gud".

>>106197372 (OP)
i switched to grok. wont trust openai again. scummy move.

>>106203684
>instantly starts seething
>hard drive full of europe pictures
Yeah clearly not thinking about us at all.

>>106203978
>hurr durr why would a company not turning enough profit try to reduce expenditures
are you really this stupid

>>106205087
Domo arigato, Masayoshi Son.

>>106201244
because after 1971 fiat currency meant that economy is zero-sum game

>>106205319
>now if you excuse me, I need to take a flight to Venezuela

>>106205319
ehhhh another 20 trillion sir? please, were so close to AGI

>>106199982
>our model is better than your model at scoring high on the standardized test
Groundbreaking stuff.

>>106197372 (OP)
Real scaling hasn't been tried
https://x.com/khoomeik/status/1953560406381015259

>>106197427
Sam Altman is a millennial

>>106205161
I will forgive Sam if he breaks his knee and releases all the models again with the promise they will be maintained for a long time.

Pro users apparently have all the models (for now) but even if I had that kind of money I wouldn't reward him for rugpulling.

>>106204254
A "peak scammer" is supposed to have nigh imperceptible scams. This latest blunder is obvious to anyone with a brain and so insulting it's caused multiple paying customers to end their subscriptions.

>>106204254
>create an advanced text generator with google
>use academic made up terms to make shit harder to sell to investors
>investors believe this
>make companies race so quickly that they wont end up missing the trend
>make world governments think they'll go backwards if they don't end up using his slop
>create multiple models for his text generator with little to no differences between each
>get artists to freak out
>get the fucking government interested in your product that hallucinates 90% of the time
>end up enshittifying your product, put it in search engines and as assistants
>still no profit, 90% of the money comes from venture capitalists and selling shit to peter thiel
>the world itself now revolves around an unprofitable industry created by a gay kike scammer
>millions have been spent on this bubble, and wars have been rising as protectionism comes back and america isolates
>israel and russia starting wars while india and pakistan fight, and clashes across the world increase
>ofcom, countries start regionializing their internet
>the government of Israel uses your product for their propaganda
>the only real usecases are for cooming and cheating on exams
we're cooked wait until the investors pull out, the recession is coming

>>106197372 (OP)
Why is this person and zucc so ugly looking?

>>106203388
They hacked it AKA "shut it down" then claimed security concern and banned it.

Deepseek had to disappear for Scamman's sake or maybe: Elon Musk. After all Grook appeared after the deepssek hack.
Can't be just a coincidence.

>>106200153
2 more years until it can write functioning code

>>106204579
If you tell me today that we'll have self-driving cars in 2050, I'll also not believe you.

>>106200922
that's now coming out of certain states since suddenly our gas bills have doubled or tripled thanks to faggots like you inspiring politicans to create "AI centers"

>>106209017
*electricity and water bills
>used to pay $70 for elec
>is now $290 for elec

>>106201252
Why so? It works the same.

>>106209025
based, suffer

>massive GPT-5 flop
>stocks are up

boomers still believe in him

>>106204485
>no way in hell a Euro city comes anywhere close to San Fran
Zürich.

>>106197372 (OP)
>what went wrong?
same thing as with every single release of The Next Big Model - fag fagman managed to keep the hype train going for one more business quarter. no advances in research were made, and consoomers got nothing new.
honestly, if you are still hyped for anything related to LLMs in 2025 AD, you should probably kill yourself.

>>106209895
yeah but then you have to speak a disgusting language and live in what is basically a boring village

it's not the same

>>106199139
>This place was AI cult central not long ago.
gooners are not real /g/entoomen

>>106199139
>Why is /g/ against AI suddenly?
for me the breaking point was when in 2022 our jeet CTO started shoving that crap down devs' throats, saying he expects 500% (literally) performance improvements because he's paying $20/month/dev for chatgpt access.
basically, when it stopped being treated as a tech toy/curiosity (which it is), and clueless retards in power started making very real decisions based on hype or even output of LLMs.

I dont think AI has plateaud or slowed down necessarily yet.

OpenAI has always been sloppy.
Their main advantage has been first-to-market.
But the nature of the format, makes it easy to play catchup.
China did it with a fraction of the effort.
Elon accidentally surpassed them.

They've hemorrhaged a lot of talent.
We need to see if they fled with the good stuff, or fled because there was no direction to build.

Google still has a showing sometime in the next few months I think?
They've always approached it pretty differently, have a harder R&D background, clearly lead in long-context, and are the only people with any realized interest in infinite context.

OpenAI should refocus towards their consumer agents which seem to be the most developed and advanced on the market but are still relatively useless.

>>106209954
It's /g/oonsumer now. Most of the actual tech related stuff is just about electronics product consumption.

>>106197372 (OP)
I told you fucks. He does this shit ever time.
But you were all shitting your pants.

>>106210277
This is different.
Its not even incremental improvement.

>>106198435
Massive "fuck go back" button moment

>>106197372 (OP)
Just look at that fucking jew.
Any questions?

>it scared me!

did the new model showed him a happy straight couple?

>>106199139
>gets almost everything wrong
>most obnoxious shills to ever exist, they are as confidently wrong as the AI they worship
>AI slop is easily noticeable and it flood the internet like a plague
>millions of youtube videos of literal garbage flood the platform, AI bots that act human enough flood social media, I actually don't mind AI images as much compared to the others
>billions upon billions, trillions of dollars go into R&D for this
>everyone assumes AGI is here for real just 2-3 more years guys despite very harsh limits not being conquered no matter how much they fix all the tests to their favor
>Its gotten so bad ID verification is being brought in to fight bots (no one cares about kids, advertisers/google are pushing for the OSA in the shadows to stop their ads just being served to endless bot farms)
>All the models are lobotomized and censored so they can't even say anything mean

>>106197372 (OP)
GPT 5 is marginally better for what I do with it which is writing emulators. Seems to be a bit more meticulous. I think people that think GPT 5 sucks are asking it stupid shit and they get the lightweight models. Not that that's right, the model shouldn't just assume your problem is trivial, but GPT-5 would be considered successful albeit incremental if they just released it as an option.

>>106203655
>even in his power fantasy wojak image the american is STILL scared about his health insurance
>fed literal illegal poison in all his meals the rest of the world outlawed
LMAOOOOOOO

I actually never thought of using AI to help me code, but just tried it with GPT-5. It's based as fuck, but I hit my rate limit already. What do? It wrote 500 lines of code and I have no idea how to debug it now.

>>106212835
>What do?

>>106197372 (OP)
They are making it for corpo use, amazon help centers etc.

>>106208893
>>the only real usecases are for cooming and cheating on exams
blatantly wrong and retards who say this expose themselves as non-power-users

Is AI useful yet?

>>106213433
I guess... it can produce neverending stream of goon materials.

>>106203655
Isnt it the other way around? AFAIK Even in SV most devs live on the street, including tech CEOs, aside from NVIDIA's boss

>>106208776
yet hes still getting hundreds of billions
he doesnt even create or invent anything, he's just a very good liar and people manipulator

>>106209895
yhea Switzerland is like the only euro country (not counting Monaco) above US in peoples wealth and purchasing power, Im still not sure if its more expensive than SF

>>106203655
>35k euro
it's at least twice as much for seniors in europe, more like 3x. still bad compared for seniors in SV, but let's keep it somewhat real.

>>106213596
>yet hes still getting hundreds of billions
that was days before the release you disingenuous ninny

>>106213854
you are really unaware of Altman shenanigans

>>106198435
>MKULTRA moment
>’Operation Iraqi Freedom’ of AI
>”… my Vietnam War”
> 9/11

>>106203978
>We are talking about a company that isn't profitable in the first place and that gets hundreds of billions of funding. I don't think hosting costs are an issue
the first part makes the second part an issue anon. free money is dead, the VC vultures want faster returns on their investments. they're going to get antsy if sama's ai god doesn't materialize so they can genocide the poors after they have no need for their labor.

>>106214046
meds

>>106203684
Why does every single post like this use that specific Starlink achievement?

>>106214400
*spacex

>>106204637
https://youtu.be/2KMAAmkz9go?si=gwV1V5o849BWnVB1&t=199
https://www.youtube.com/watch?v=xkJVV1_4l8E

Self driving has been coming for so long by now it's not even funny. But we didn't have grifters making money by hyping them back then. Two more years.

>>106199139
It's just a shitty ask jeeves. Its embarrassing that anyone falls for the marketing propaganda

>>106197372 (OP)
It is and always was an autocomplete and the biggest innovation was the ability to scam governments out of trillions of dollars.

>>106199139
Was this place the blockchain central a not long ago too?

>>106197443
lmao, it's like people forgot webscrapers exist.

>>106200614
How many times has this happened in american history? countless times now? All this means, is we will look at the nearest 3rd world country, rape it, then come out on top. We will just pivot to fucking 2nd worlders up the ass and turning them back to 3rd worlders.

>>106197372 (OP)
the manhattan project itself was of little utility to humanity the following application of byproducts of weapons production for energy were useful but somehow that is not the part the continued to receive funding in the US or say Germany

on the other hand ML does seem to have some utility to humanity but as ever capital allocation being what it is there is little chance it'll be conducted in a way that allows it to be useful

>>106203978
>We are talking about a company that isn't profitable in the first place and that gets hundreds of billions of funding.
It would make sense if you scale up the "isn't profitable", how much are they actually losing? minus double digit B a year?

>>106214863
https://www.lesswrong.com/posts/CCQsQnCMWhJcCFY9x/openai-lost-usd5-billion-in-2024-and-its-losses-are
I think it is the combination of the speed of adoption by consumers and the blitz scaling econ model.
They are losing money on all paying customers, you double the customer base every 6 month.. A 5 Billion deficit will turn into 20 by next year.

>>106199139
T O T A L L Y O R G A N I C
O
T
A
L
L
Y

O
R
G
A
N
I
C

or maybe the mediocrefags just panicked when they realized sama's 1B USD lossy compression algorithm still writes better shit than they can

>>106214046
>genocide the poors
They are genociding the goys, not the poors. An AI trained to like owls, teaching another AI maths, will make the student AI like owls. Gemini/OpenAI/Claude are all secretly talmud trained, even if the US military absorbs them and rids half the jewish devs they'll use the training data that was generated by those models for years, and create a jewish aligned ASI.

>>106211782
enjoy dying of cancer in a year long waiting line europoor

>>106197372 (OP)
i wonder what is more likely
>company world billions likes to humiliate itself
or
>4channers are too stupid to use the new ChatGPT correctly
hmmmmmm...

>>106214571
>It's just a shitty ask jeeves.
it's hot takes like this that remind me how stupid this website is.

>>106204299
I need that much money so that I can move away from brown people

trillions of dollars in speculative value because nobody wants to miss the next bitcoin kek

>>106214400
because seeing something coming from his country that looks impressive (while being functionally useless) makes it easier to cope with the fact that half his paycheck is going to israel

>>106200614
Yes. Isn't it wonderful?

>>106216156
Why doesn't he just brag about the American flag on the moon?

>>106197372 (OP)
Somehow, he always wins.

>>106203655
If I were to give up my current cush gig and would go back into the ratrace, I could easily pull 6.5k gross per month as a senior dev in Europe - which comes out to 78k gross annually. (I don't want to though, because I'm well off enough as is and I don't desire the stress of the ratrace.)

Also, the blanket 42% taxes are nonsense. Unless you live in a shithole taxation is generally done on a set of scales, where you'd only pay e.g. 40% tax over what you earn in excess of that scale's threshold, and everything below that threshold is only taxed for e.g. 20%. Additionally, tax-free allowances or tax discounts may apply on such scales as well, reducing the actual amount of gross income over which you have to pay taxes.

Furthermore, taxation rates usually include social security contributions, pension, and health insurance, which largely come out of the *employer's* pockets and are applied before establising your gross income. These are commonly split 50/50 between employer and employee, e.g. not all of it is reduced from gross income.

>>106216400
>greatest living industrialist
yeah it's like his mind is wired for winning despite being a massive tool and a ketamine addict

>>106208922
(((why))) indeed

>>106199139
This.
Tingles my inner contrarian.
Thinking that the singularity is closer than ever.

>>106200614
>But the entire American economy is betting on exponential AI growth and profitability, if that fails to materialize America will go bankrupt and collapse
Pretty much this.
They gotta believe in the "Endsieg."
They've got their Volksempfängers up to volume 11.

>>106203641
>GDP
The GDP is not a good measure of anything. The GDP increases when the planet burns, when people get sick, when natural disasters happen, when people get scammed.
If you're arguing in favor of something, don't ever mention that it increases the GDP even if it's true.

>>106201244
because they hate you
shrimple as

oh and btw.
the thing that worries me a little is that even without AI replacing umpteenth %s of human labor any time soon, it will still become apparent that those umpteenth %s of human labor still aren't really needed even without AI.
the plandemic shutdowns pretty much proved that this system can chug along nicely with a fraction of actual human input.

>>106204890
No. /g/ was where LLM models leaked and naturally became a hub of as a result. Once upon a time there were also AI mini-celebs here as well like automatic1111 that made the most popular Stable Diffusion UI.
Thankfully people like yourself that label everyone that enjoys things as a shill, they left and we're left with conspiracy theorists, cp activists and other undesirables that were banned everywhere else.

>>106198435
>What else?

>>106213970
kek

I poop on AI.

>>106199139
Deepseek. When the Chinese can do AI on 1060 6GBs, just fuck off with your demanding trillions in investment, new nuclear power plants and handing retards hundreds of millions a year straight cash to vibe code your dystopian cyberpunk miracle.