Thread 512353417

I was a kid when the dot com bubble burst, I was at uni during the 2008/09 financial crisis and I was in the process of career plateauing in the 2020/21 covid crisis (which pushed my career upwards).

It’s current year and fucking AI looks to burst like the bubble it is. AI is currently getting partially worse. Sure, it is helpful for supplementing Google and getting certain insights, but the slow progress in recent times on AI is reeeeally worrisome.

>>512353417 (OP)
>AI is currently getting partially worse.
No it's not doomer shitskin. There has never been a tool in human history that has ever even come close to improving at the rate of AI, OpenAI just released an open source model TODAY that beats their premium top tier model (o3) from less than 6 months ago. When was the last time you saw a technology that went from state of the art cutting edge to FREE and you can run it on your laptop locally in 6 months?

>>512353523
>thing people are working to improve gets better
So?

>>512353589
What are (You) working to improve in your life...and is it getting better?

>>512353417 (OP)
The AI explosion isn’t even close to where it’s going to end up.

>>512353417 (OP)
AI has been controlling financial markets since the dot com bubble. I'm tired of brainlets coming in late trying to cope with denial.

>>512353523
It is getting worse if you actually use it daily. I use different models, and the restrictions built into answers have heavily increased, so it takes a lot longer to get an answer compared to just 6 months ago… not to mention that hallucinations have gone up and have become more sophisticated.

That's what they planned from the beginning
There were AI chatbots 30 years ago which worked almost like they do now

>>512353724
>It is getting worse if you actually use it daily
No, it's definitely not.
>so it takes a lot longer to get an answer compared to just 6 months ago
No, it doesn't do that either unless you are using a thinking model and giving it a hard or agentic task.

>>512353768
>There were AI chatbots 30 years ago which worked almost like they do now
Yeah, there were totally AI chatbots in 1995 that could answer any question and win a gold medal at the IMO. It was just the dial up internet that held them back.

>>512353798
We use AI for financial and technical modeling and it has consistently gotten worse in the past 6-9 months. We still need to use our old pre-AI methods and can just use AI to double check and get solution leads - the amount of time chasing hallucinations has also more than doubled. We use Copilot which is OpenAI based.

>>512353840
>n 1966, the world’s first chatterbot program - a computer program designed to interact with people by simulating human conversation - was created at the Artificial Intelligence (AI) laboratory at the Massachusetts Institute of Technology (MIT) by Joseph Weizenbaum.

>https://www.botsplash.com/post/chatbots-a-brief-history

>win a gold medal
Wonder who awards those medals...

>>512353417 (OP)
>It’s current year and fucking AI looks to burst like the bubble it is. AI is currently getting partially worse.
Hopefully soon. Regulate """tech companies""" like Google, Microsoft, Amazon, Meta, Nvidia and other crap like that. They are causing brainrot. They are monopolies. Send them back to USA. Make forums and imageboards great again. I think AI stuff are good, or they can be good, like machine learning or even llm that are truly open, uncensored and enhanced with RAG. But not in the hands of """tech bros""" or their indian lackeys.

Here's the main parallel I see: The dotcom bubble happened because there was a lot of hype surrounding what the internet could be. BUT, the use cases weren't there yet, it needed time to catch up, hence the bubble popping. I see the same thing with AI. AI is getting better and better every single day, and it seems almost weekly that AI makes significant jumps in certain fields. HOWEVER, the case uses have not caught up to the actual progress. The progress is going by TOO FAST actually, the economy can't actually make use of half of this shit yet in a meaningful way. Even worse, while a lot of it is very impressive, it's not "there" yet for a lot of actual, consistent commercial uses, yet with how big the AI market has grown, it's ACTING as if it IS used everywhere.

So yes, I do think there will be a huge bubble pop.

>>512353589
Muh ai allows automating automating tasks. Imagine your job is to automate a certain task or function like building a cotton gin. Now imagine you can ask muh ai to automate not only building the cotton gin but coming up with how to in the first place including procuring investment funds to do so as well as novel ways to do it better than a cotton gin. In seconds.

>B-b-but it can't do 100% of all tasks perfectly yet tho!

>>512354107
AI is literally leveraged to manipulate markets by various actors (ai agents).

>>512353417 (OP)
>kikes invested all that money into AI so i could make up pictures of rebecca cunningham from disney's talespin re-enacting events in jewish history where God punishes them
>and if this service fails
>it "could"
>"take the entire economy with it"

LOL fucking DO IT

>>512353972
I'm just glad you're here instead of working in the drone or shell factory babushka.

ofcourse the lyin' TREASONOUS DEMONCRATS want the BASED AI companies to collapse
sick of these WOKE communists

>>512353417 (OP)
Kek I will laugh so hard if the AI bubble becomes the next Dot Com shitshow and is used as a distraction from the Epstein files. Prepare to see so many tiktok zoomies, streamers and retards on the unemployment lines.

>>512353417 (OP)
The difference is AI shows genuine promise. Two models just got gold medals at a math competition recently. Google just came out with a World Simulation model. The problem is that the world can't figure out what to do with all this fast enough. This is drastically different from the dot com bubble where people thought Pets.com was going to be huge just because it was a website. And certainly it's nothing like the crypto/nft bubble as that's all useless.

>>512353417 (OP)
Also the companies pouring shit tons into AI are mostly making billions of revenue already. They can afford it. The money isn't borrowed, it's revenue. This is true of Google, Microsoft, and Meta at least.

>>512355860
lol I wonder why they always report revenue and not profit.
They're not profitable, except the shovel seller (Nvidia)

>>512353417 (OP)
funny thing thats happening. Friend of mine is about to be let go from a luxembourg company. he works as a financial data analyst and his job is 100% replaceable by AI. His higherups are now banking on AI taking over all the Fond management und thus dont need any humans anymore. If the Bubble pops and that company will need real data analysts again that can properly research and not hallucinate data, he will be in dubai. Basically if the Bubble pops, alot of euro companies gonna suffer because they are led by 70IQ BWL idiots

>>512355820
Proof of work crypto and ai have the same inputs.

>nfts
No one cares.

>>512353417 (OP)
Oy vey goyim two more lines of code for sentience just you wait

>>512355952
>If the Bubble pops and that company will need real data analysts again that can properly research and not hallucinate data, he will be in dubai.

Your friend can't read hundred of different markets simultaneously perform pattern recognition on them, and keep up with keywords in contemporary news via the internet in real time.

>>512353840
>>512353972
>>512355820
>Yeah, there were totally AI chatbots in 1995 that could answer any question and win a gold medal at the IMO.
>Two models just got gold medals at a math competition recently.
They didn't win gold medals.

>>512353523
>model designed to do well at a set of tasks we made up and tweaked to show off how good our model is to VCtards is really good actually!
>no ignore real world use being completely fucked and worthless look at the METRICS
>PLEASE GIVE US MORE MONEY THESE CARDS ARE A HUNDRED GRAND A PIECE MR. NVIDIA WON'T STOP RAPING ME

>>512356168
What's the point in doing that if same news were AI written too?

>>512356249
Pattern recognition.

>>512354107
>Here's the main parallel I see: The dotcom bubble happened because there was a lot of hype surrounding what the internet could be. BUT, the use cases weren't there yet, it needed time to catch up, hence the bubble popping. I see the same thing with AI. AI is getting better and better every single day, and it seems almost weekly that AI makes significant jumps in certain fields. HOWEVER, the case uses have not caught up to the actual progress.
It's the exact opposite. The use cases are easily understood, but the capabilities simply aren't there, despite the hype. Also, one key difference between the dotcom era and the LLM era is that hardware isn't advancing at the pace it used to back then.

It was never inflated to begin with. Corpos keep trying but everyone sees through the bullshit

AI is pretty good at making music and racist videos portraying black women accurately.

>>512356397
>4 gorillion novideo market cap
>not inflated

>>512356349
Patterns AI halucinated?
there's no point in analyzing anything stocks because you can't predict human behaviour. This has been proven hundreds of times. There is a point however to be friends with business and get insider info beforehand, but you don't need AI for that

>>512356249
Pattern recognition on a level humans and eventually humanity as a whole wouldn't be privy to.

>>512356238
qq faggot

>>512356239
Yeah and the set of tasks AI is great at just keeps expanding and expanding while your set of tasks just stays the same.

>>512356349
>okay, let's break it down. here are the pattern's I've found in the data you've told me to analyse
>what the fuck this is a bunch of incoherent gibberish you've just made up
>you're completely right! let's break it down. I was only giving you that response as an example of a hypothetical pattern analysis

>>512353417 (OP)
It's not going to burst, it's not getting worse, you're a retarded nigger and they're going to kill you first.

>>512356427
He is a retard, fren, don't mind him, it's obviously that all AI bubble is collapsing.

>>512354107
AI still cant actually be used to remove human devs for creating agame, right? If so, then your theory is 100% right.

>>512356427
>Predicting human behavior
Manipulating human behavior.

>>512353523
> uhm goy, did you know OpenJewI just RELASED A NEW MODEL?
> world-shaking news, goy!
> buy your subscription now!

>>512353417 (OP)
What the fuck is this idiot logic? if AI popped tomorrow then the economy would take a hit sure but we've already got systems in place like we did before it so things would truck along without much issue. Just sounds like some scare tactic to keep funneling money into it.

>>512353798
> uhm akshually no, goy
> trust be benchmarks, ignore your eyes

>>512356445
They didn't officially participate, you fucking retard.
Anybody can solve the questions and claim the gold medal, but that's simply not the same.

>>512353417 (OP)
Dotcom was about the rowdy VCs investing in everything with the right buzzwords, coupled with mortgages to niggers.
AI is about corporate tech sponsored by (((our friends))) and as such will be regulated to hell and back and will be successfully used to guard your sorry ass forever and ever.
Zoomers are genuinely retarded.

>>512356454
yeah now it can pretend to be capable of performing a whole range of tasks while shitting out utter cunt gibberish! literally indistinguishable from fucking jeets! how good is AI????

The moment that people will ignore products with AI, disregard AI talk etc.

>>512353724
>not to mention that hallucinations have gone up and have become more sophisticated
Hallucinations are caused by 2 factors:
1. Model inbreeding which has become unavoidable due to the training data being contaminated with AIslop
2. Lowering model precision (FP16 to FP8/FP4) to cut costs

>>512353724
Its prose got better and hallucinations minimized but it's as useless as in the past. Also fuck its "creativity", oh you wanted two buttons? Well I've made one red and one blue and while I was at it I've hidden them inside another menu just because.

>>512356562
To be fair it was a stock market driven bubble, this one is venture capital driven, it's more isolated.
Also dotcom was masterfully burst by the FED it was a soft landing.

>>512353962
is picrel related or unrelated?

but you're right, i wouldnt trust any openai model with anything that truly matters. it does hallucinate and it purposely gives you red herrings on political and woke stuff. I mainly use it to learn and in that regard its a heaven send. It's like having the best university professor on any subject at your desposal 24/7 and he wont give you attitude for asking him to explain a second time. the limits, however, is that cutting edge stuff it wont know and even using it to learn, it will make stuff up. and if you suspect its wrong it can take 15 minutes to quiz out the error. so you cant reliably trust it for anything. if you use it to do anything that matters, having to double check it makes it cost prohibitive or redundant

>>512353523
>New models still fuck up hands.
>B-but I saw a clip where the hands are fine!
Yeah, because people use motion cap or separate 3D-model plugins to get around the issue.
The biggest issue for AI is not AI slowing down, but investor confidence evaporating thanks to pajeet scams being exposed like Builder.AI and whole Deepseek announcement making investors question if theres any actual money to be made with AI if everyone can access it cheaply.

>>512353523

You can have a bubble (and often do) even if the tech is useful and raises productivity. That's what happened with the internet and dot.com bubble, a shit ton of startups sprung and investors began pouring money on every dumb idea left and right hoping to nail the next Google or Amazon, getting into debt for it etc.

So when it became obvious that a lot of those companies are worthless the bubble burst and panic ensued. Even libertarian economies agree that busts like that are constant in capitalism and consider them good for "correcting" bad investments and flushing out bad companies

With AI, you have an even bigger problem of boomer investors not understanding the technology and AI companies overselling pattern recognition and creation as high intelligence which it isn't. Improvements in artificial intelligence may come but currently all the improvements are horizontal, from more and more data / computing power. To put it differently, we're not developing AI from a bug to dog intelligence, we're just using more bugs

Which is a big problem because boomers are investing en masse and believing it's a technology to replace human intelligence workers everywhere instead of being a tool to make them more productive through automatization of specific, basic cognitive processes. That is a massive misinvestment so the wave of "corrections" is inevitable and will be a fucking cunami on an already insanely fragile and strained global economy

>>512353417 (OP)
SAVE US TANGERMANN

>>512356646
>>512356692
>>512356748

It's like the tulip bubble. You know fundamentally there's no use case and no problem being solved by tulips but people project their phantasies of becoming rich before the bubble inevitably bursts.

>>512354210
Nobody is using AI for that shit, market bots are simple if-then scripts based on complex financial models that don't leverage "AI" or Nvidia cards.

>>512356798
>The biggest issue for AI is not AI slowing down, but investor confidence evaporating thanks to pajeet scams being exposed like Builder.AI
The reason those pajeet scams exist in the first place is because AI isn't progressing at the rate they are claiming. Remember the Amazon self-chekout stores? Amazon's goal was for purchases to be at least 95% automated and 5% pajeet-driven, but it was nowhere near close and ended up on a 30/70 split, which is why it got shut down.

>>512356646
Yeah, we could beat the jew in 10 minutes, all it takes is an emergent NPC sentience.
Two more bubblebursts.

>>512356748
It's not as VC driven as the dotcom was. I don't see many garage firms making their own AIs, not when compared to dotcom.
There will never be a real crash, because that would endanger TPTB. If anything, it will be slow and steady, resulting in pods, no personal property and plenty of chemical happiness.
And AI is great at creating happy slaves, ergo it will never go away if (((they))) can help it.

>letting coomers make endless pictures of sonic birthing and pajeetbotacounts elon musk pepes for free isnt making any profit
No shit. Twitter is now 75% bots, its AIs trying to sell other AIs crypto.
Thats why this crap gets astroturfed to hell, all it has is GROWTH but no substance, and nothing showed this better then the chinkoids throwing their modell on the market actually open source just to tank Nvidia stock

>>512354371
Dump folder plz, my dick can only get so hard.

>>512356845
To be fair the industry is not ignored in America, sooner you'd notice that Europe forgotten how to grow compared to America with all its problems.
https://youtu.be/O1dbtWZFsIk?t=422
And it's like that one then it's a nothingburger.

>>512356918
Sure. Kids in their basement use it like that. The groups building literal powerplants for their datacenters are literally using ai to manipulate various markets and groups of people associated with them. Zoom in.

>>512355820
>Two models just got gold medals at a math competition recently.
Yes, goy. Trust the benchmark data released by literal Jews.
Jews would *never* lie or game the system!

>>512357058
Now generative image AI is utterly useless, the main economic driver is the LLM and with its capability to give or take act like an autonomous brain.

There's no reason to believe that this can't be productive on its own so GDP growing, now the question is if it's sustainable long term, especially in more meaningful areas like programming where it would incur astronomical rewriting costs and introduce bugs and logical errors that will never be gone.
If we're not Marxists about it and deny that machines (in this case software) cannot generate value then yea it's a destructive giant bubble but if it adds value, which it obviously adds then we should downplay its scale.

>>512356918
"AI" models need to not only have an edge, but pay their training cost before the edge collapses.

Good fucking luck to the people trying that shit.

> porn for free isn't making any profit
Not everything is about profit, goyim.
Well, direct profit anyway.
We can print your "money".

>>512353417 (OP)
I hope it bursts and disappears already

>>512353417 (OP)
It isn't a bubble, not yet at least. We are looking at computer systems that render huge segments of the workforce obsolete. Many white collar jobs are teetering on the edge of collapse.

>>512353417 (OP)
I'm looking at the author, skimming the headline, trusting my suspicion it's jewish, and moseying the fuck on out of heeaarrr
*whip crack*
HAIYAA

>>512357379
>We are looking at computer systems that render huge segments of the workforce obsolete.
Booting some graphics dudes and service desk isnt exactly "huge". Corpos keep buying this bullshit and integrating AI into any new app but is it even useful? I still have my doubts
t. dev

>>512353417 (OP)
We're long due for a new financial crisis. COVID and the massive quantitative easing canceled it for a while, only to make it worse in the future. The AI bubble popping could be what starts the fire, maybe it will be something else, but all that fake money will soon be valued to its true value, which is near zero. Something will give in very soon, and the states won't be able to bail the failing banks for very long. 2008 will look like a nothingburger in comparison.

>>512357174
>There's no reason to believe that this can't be productive on its own so GDP growing
In theory, it would have a deflationary impact on services, which are ~70% of the GDP of advanced economies. Similar to how people aren't buying $500 encyclopedias anymore when they have access to kikepedia for free.

>>512357735
This, but it won't be AI bubble bursting, at least not in the 2008 sense.
Mainstream crypto will be wiped out. Personal savings will be "lost" or requisited. Internet will be fractured and posting on social media or accessing an "adult" side will require ID.
There will be another war in Europe.

>>512353523
>Beats their model
Ok but does it do anything? This is literally like all the dorcom bullshit years ago: “at this internet page is better, invest in it” and then you find out you are paying for a site to be up and be looked at.
You AIfaggots are telling me we will have robots replacing us for shit works for 5 years. I open any fucking reciept and I still see “we need people who want to do shit jobs”.
AI is on the same level of “buy my coun we are decentralizing the economy, ignore the volatility doomers”. When the bubble pops and the governments will use another “economic crisis” to fuck you all over, I will suffer it too but I will suffer it smiling at your pain.

>>512357895
I believe the recent push for censorship and population control is linked to this. They're preparing for the next big one. They know it will trigger huge revolts because when everything is centralized, everything crumbles all at once and a lot of people will starve pretty quick, so they need to keep control with rationing and a police state. But it only works if most of the institutions stay functional, which probably won't happen if the system falls USSR style. Sizing bank accounts will maybe happen, but it will be useless since the FIAT money won't be worth anything anymore. And yes a future big war will probably be the end result, that's the modern jubilee. I hope I'm mistaken but we're closing to the 100th anniversary of the 1929 crisis.

>>512358199 checked
Anon, IT companies are now encouraging use of AI. One of companies I work for now has its own sanitized instance running on premises.
I've seen emails written by AI and sent unironically to the management.
I've even seen some of my juniors trying to vibecode.
I'm not saying it is completely usable for everything (it isn't). But it helps retards to feel less retarded, so it will stay and it will be used more and more.
Here, a SK minister used an AI to write her a speech, which somewhat inappropriately quoted Bram Stoker ... https://www.ta3.com/clanok/958833/krv-je-zivot-vyhlasila-ministerka-dolinkova-na-vyroci-narodnej-transfuznej-sluzby-citovala-drakulu
Can't stop the "progress".

>>512358354
I still don't think that the next manufactured global crisis will lead to governmental collapse, for the exactly same reason why nukes won't be used. Crises like that are eugenic. (((They))) don't want that.
Besides, USSR collapsed for various reasons, but it didn't have to for another few decades.
However, jews wanted their privatisation cut and it was dangerous to keep two systems alive - people mustn't have the possibility of comparing their lot to anything else.

>>512353523
>OpenAI just released an open source model TODAY
Yaddayaddayadda.
They are building data centers and need new power plants for them in order to run AI models that are just as useless as a random pajeet IT peon, but at least they doesn't smell like shit and curry.
Meanwhile internet search engines are getting useless due the AI slop everywhere and the stupidest decision makers are start to listen to the AI models in their decision making process.
It won't be just a bubble burst, itt will be a full scale meltdown. Hopefully there will be cheap used video cards on the market after that.

>>512358199
You're a retarded emotional child. Yes, AI has solved some of the hardest and most important problems already (see Demis Hassabis, John Jumper and David Baker nobel prize), and it has just begun. There is nothing stopping AI from replacing everything, literally.

>>512353417 (OP)
Remember to buy when there's blood on the streets.

Everywhere I see it there are tons of people calling it slop (it is). The only companies and people using it now are out of touch (mostly boomers).

It’s already mostly bursted visually. ChatGPT has uses but you can always tell when something has been written by it . It’s information isn’t always trustworthy as well

>>512353417 (OP)
im still waiting for the internet bubble to pop because when that happens well that's going to be real yeah it's gonna be somethin'

>>512353417 (OP)
There might be side-bubbles, but the main thrust is no bubble. Check out Genie 3 and get some sense. This is a massive breakthrough that allows for older tech to gain new function, it will keep growing. Might as well call agriculture or iron rail a bubble.

>>512353523
You’re retarded. LLMs are not “AI”

>>512353417 (OP)
Release them!

>>512353417 (OP)
I don't give a shit. If it bursts, then we will survive as we always have.
Kikes won't, tho. And, that's a good thing.

>>512358593
I believe that the next big war won't be a World War in the traditional sense, but a global civil war. It has already started, incessant division within our countries will ultimately lead to this.

>>512357174
>act like an autonomous brain
Act like an autonomous parrot.
There is zero thinking involved in what an LLM does.
It's basically a slightly different search engine.

>>512354371
Kek that show was gold

>>512358948
Ok very something analogous to what might go in our brains, due to comparable outcomes.
This was not a talk about consciousness and irreducibility.
A factory machine doesn't have to have qualia and intentionality to produce commodities, now the interesting part is that these stochastic parrots, which are black boxes largely can reproduce our intentionality in meaningful ways often following rational patterns and closely following the intentionality of our inputs as well.

>>512358635
>see Demis Hassabis, John Jumper and David Baker nobel prize
That's ML, not AI, you moronic neanderthal ape.

>>512353523
>>512353417 (OP)
AI was shown to us. We saw the potential in its ability to predict, build, and organize, etc... then we saw it thourly lobotomized or gate kept. The only part of AI that I will enjoy is watching all those retards with tech/data driven jobs who told me that automation will bring us to Utopia end up as poor, jobless, and betrayed.

>>512353417 (OP)
It's not a bubble.
I'm a Senior Programmer.
AI will easily be 3x better in 5 years guaranteed if not more. The agenda is we will put AI in charge of as many world systems as possible. Even if it is wrong <5% of the time that's the same as human error in the same position. You will be judged by AI Judges and AI Juries, ticketed by AI Police, etc. I didn't say I like it, but I know that's what will happen.

>>512359199
In otherwords, a fully AI driven world. AI will govern the goyim as a proxy for the Elites so that they don't have to try so hard.

>>512356692
You are not being explicit enough in your prompts

>>512359257
No u.

>>512353523
>https://gpt-oss.com/
it's shit, AI will never be anything but an assistance, it's almost saturated now

>>512359033
They're not stochasitc anything. They solve novel math problems. They're not black boxes, either. You can peer inside and see the labeled regions light up.

>>512353523
Actual AI adoption to replace labor and make things cheaper is really hard. Big tech sells hype drivel
https://youtu.be/H6jlGChbV_w?si=zbPvYvJlqb1YjGIt

>>512358707
>This is Jewish megacorp ad a massive breakthrough
Fucking retard.

>>512359033
>Ok very something analogous to what might go in our brains
If by "our" you mean Romanians and other gypsoids, then yes.
Human brains work differently though.

>>512353417 (OP)
Nope. Zou vil get replaced. Corporate will make sure it succeed and they have the means. the savings are Too juicy to let this chance pass.

>>512359066
Machine learning is AI.

>>512359387
>They solve novel math problems
They do not do that no. They might employ mathmathica syntax in a clever way that you probably wasn't aware they got from some hidden page on the internet but its core model cannot even add two numbers consistently.

>>512359716
I accept your concession.

>>512359199
Then what is the point of keeping humans around once everything gets automated? It will create a global existential crisis.

>>512356461
>they're going to kill you first
Doubt it, the AI can't even say nigger without getting it's plug pulled.

>>512353417 (OP)
No anon Nvidia really is worth 4 trillion

>>512356822
High IQ post.

>>512359818
By "AI" everyone means LLMs and other generative neural networks, you duplicitous ape.

The ML used in computational chemistry has nothing at all in common with OpenAI's grift and has existed for decades before ChatGPT.

>>512359874
Only niggers have a stochastic parrot running their brain.
Humans work differently.

>>512356456
eventually this Ai bullshit will costs 100$ per promt and stupid fucking idiot jeet ceos will scramble to get human analysts back into the office because they are cheaper and work more reliably than an "AI" LLM. question is, if these new hires are as reliable as the workforce you initially let go at the start of this stupid craze

>>512360008
all this shit is, is a glorified data crawler with a better search engine than google and jeet CEOs treat it like the second coming of the computer christ.

>>512353417 (OP)
from what I see with my eyes it just accelerates faster and faster. It is filling niches that didn't exist and for which there's nothing to go back to

>>512357228
You have to understand printing money doesn't help in the long run, at that point you might as well give away monopoly money because it will be worth more then real money

>>512353417 (OP)
It will burst when they're ready for a full reveal of the real function of """AI""". Until then, they'll continue to string useless eaters along with promise of AI utopia and scare tales of AI run amok.

>>512360209
>from what I see with my eyes it just accelerates faster and faster
Anon, this is called schizophrenia. The rate of discernible improvement in AI consumer products has essentially gone down to 0. All you get now is corporate marketers announcing that AGI is coming in Two More Weeks over and over.

>>512359033
How many times have you had sex?

>>512360399 checked
Doesn't help who? You and me? Definitely, fiat isn't money even by the mainstream definition. Inflation is hidden taxation, goyim should care about what to eat instead of meddling in politics etc. etc.
Does it help (((them)))? Of course it does. It harms us, that's enough for them. Besides, they own everything thanks to trickle down.
Look up why pajeets own so much motels in your country. Where do you think these money came from?
> printing money will lead to DOTR
Yeah. No. Not with AI, communication control etc. - all those things which are so wonderfully under way already.
Sorry for the blackpill.

>>512353417 (OP)
When the Ai bubble pops the stock market will crash but it won’t be as bad as the dot com bubble since Ai at its worse is still useful
It’s just not as useful as big tech scammers pretend

>>512353523
>omg ChatAI just released a model that scored 2% higher flogblops per pemblort than AIChat!!!

>>512353523
lmao you're a fucking gullible retard

>>512360008
>By "AI" everyone means
I don't really care what retards mean when they use words incorrectly. AI is not limited to LLMs. Not even ChatGPT is limited to LLM.
>The ML used in computational chemistry has nothing at all in common with OpenAI's grift
Nothing in common? Both are AI, both utilize similr underlying principles of neural network/statistical modeling to analyze patterns and make predictions.

>>512353962
>We use copilot which is openAI based

Thats your problem right there. You dont use copilot for accuracy tasks, you use it for work tasks, creative tasks, compiling or art tasks.

Perplexity is the one you use for tasks which need consistently accurate, well sourced information.

You CAN use open AI with accuracy, but you MUST use 4.5 with research mode. Then it behaves a lot like perplexity.

You have to know which models excel at what in order to use it accurately. I am in industrial real estate and the amount of time it saves me on research of markets, build projects, local economies, etc is just amazing. I get 4x the work done

>>512353417 (OP)
The AI bubble isn't a bunch of startups doing IPOs for ridiculous valuations though, it's yugely profitable companies plowing their profits into buying up as much silicon from Nvidia as possible.
All the yuge numbers being thrown around on AI datacentres is for the Nvidia gear inside them. The cost of constructing the buildings and running them is chump change in comparison.
The money from selling access to this shit is really starting to take off now as well.

>>512360675
Agenda 2030 is about gaining complete control of the population before FIAT money implodes.

AI tech bubble is probably bigger than dotcom on a percentage basis, in dotcom it was new corporations playing with VC money, now it is trillion dollar established corporations playing with total stock market scale money

>>512361418
I don't think it will implode, TPTB have no incentive to make it unusable. They have absolute control over it.
There are no organized good guys fighting for freedom in our countries anymore, therefore no one has the incentive and the power to harm petrodollar or any other major currency.
Atenda 2030 will be implemented, one way or the other.
There is an alternative scenario requiring economic and social destruction of USA, but ... we'll see.

>>512353417 (OP)
good,
fuck the fake economy,
fuck chatgaypeepee
fuck Anti-Intelligence
fuck kikes

Re AI, read the transcript to this shit. I heard it Grimerica Outlawed and it was a bit unsettling or it just went off the rails. Lots of DAN vibes.

https://substack.com/home/post/p-169093845

>>512360996
All you're doing is to reaffirm that the word """AI""" doesn't mean anything anymore. It is whatever your corporate marketers and jewish academics want it to be on a given week.

>>512353672
It's hopefully going to end up at vidrel.
The only hope the human race has from getting out from under (((them))) is TKD, and the only hope for TKD is nuclear Armageddon.

>>512353417 (OP)
AI hype is an IQ test. Anyone convinced that these overbaked text prediction models are going to completely revolutionize the world has an IQ of 90 or below.

They just released an open model that is as good as the real deal that can be fun in a gaming computer. It's not bursting anon.
>>512353724
Then you should be excited for open models, not saying "it's over"

>>512353962
>AI in the (((stock))) market gets worse as the AI improves
>using logic models to map out insanity
>expecting magic, instead of predictive reasoning

Yeah, can't imagine why that's not working out for you.
Hey, now good news. If AI can't even make sense of it, you know total collapse has got to be close.

>>512353417 (OP)
first mass casualties that cannot be burried by lie media

>>512362283
>AI hype is an IQ test. Anyone convinced that these overbaked text prediction models are going to completely revolutionize the world has an IQ of 90 or below.
It's the intellectual equivalent of the coof psyop.

>>512361020
This. I don't think people realize how fast this field is accelerating. Most people still think gen = chatgpt, not realizing it's one tool in a sea of emerging tech

>>512353417 (OP)
>ahhhh It’s 1901 year and the fucking Nuclear Fission speculation looks to burst like the bubble it is. Raduation is currently getting partially worse, marie curie fucking fried herself. Sure, it is helpful for bouncing dots on a screen around and getting certain insights, but the slow progress in recent times on radiation is reeeeally worrisome.

T. Heisenberg

>>512362396
In your next post, provide some verifiable products that demonstrate this "acceleration". Notice how your blood pressure is rising as you find yourself forced to reach for the statements of jewish CEOs and Jewish academics instead of actual, usable products.

The entire economy is based on expanding consumer debt worldwide, backed by expanding pools of fiat currency, underpinned by oil extraction. Ha, yeah, llms are sooo important to our economy.

>>512353417 (OP)
Thank fuck for that

>AI and its energy consumption will become a catalyst for technological advancement and the emergence of new energy sources - perhaps something like a "quantum" or "stellar" reactor. What used to be sci-fi dreams about flying to other galaxies is now turning into a real, present-day need - a demand for large energy here and now.

I was told this today by an ai fan boy. This is not the first one that has somehow pivoted ai into unlocking unlimited power sources so I guess it is the narrative somewhere. The other guy that was keen on this idea posited new energy sources would be unlocked by nation states in order to compete in the ai race. It's honestly more insane than crypto bros

>>512353417 (OP)
When AI exponentially grows and when you're lying on your deathbed wondering what comes next, and when they're rolling out brain/memory archiving in the next 40 years... you'll understand.

Also, OP is a fag and should not be archived.

>>512358199
AI is replacing low to middle management, programming, and all things botfarm, propaganda, 'journalism' and shilling in the next 2 years you won't be able to get a job in those areas, because why pay someone to do a shirty job?
Even if they're good at it, their production value will be worthless.
85% of all employment is going to evaporate in the next 10-20 years.
Good job their doing something about the 85% of the human race that is about to become redundant, and already succinctly proved how worthless they are

>>512362660
>When AI exponentially grows
How many more two-week spans?

>>512353523
>Run it locally

Kek, easiest tell that this is prime bait.

>>512362661
>AI is replacing
In your next post, provide objective evidence of """AI""" replacing anyone on any meaningful scale. You won't.

>>512361796
>They have absolute control over it.
I think that this statement is wrong. Sure they can print as much as they want, but they can't prevent it from becoming worthless. On the contrary they're only accelerating that process. At some point the system will become unsustainable and something will give in. Or maybe they'll manage complete societal control before this happens, but then we're fucked for a very long time.
>There is an alternative scenario requiring economic and social destruction of USA
The worst that could happen would be secession and potential civil war. This is what I was talking about in my previous post, global civil war. It would spread to the other US aligned countries because of the same causes, that lead to the same consequences. Not very likely for the moment but who knows.

>>512356168
But there's then no competitive advantage there, because everyone can then do that. It ends up like High Frequency Trading, where finance cos were seeking to move their data centers as near to NYSE as possible to maybe shave a millisecond off the reaction time. And then the NYSE ended up having to build in trading safeguards because all the HFT shit was causing flash crashes everywhere.
Now if you imagine this extrapolated, AI shit being let loose not only on safety railed stock trading but on every imaginable financial market or derivative, acting autonomously in a million different directions simultaneously at a speed that's impossible to stop.. wut go rong indeed. The 'AI bubble' imploding is not the real danger of this shit, rather the inevitable casualties when it explodes.

>>512362694
I'm gonna be honest with you.
About twenty years.
Computers twenty years ago were primitive compared to the ones now. Now, they double themselves every year or so. But that's still not enough for full mind archiving.

>>512353417 (OP)
>ai bubble can burst the entire economy.
who even uses this shit?

>>512362797
>About twenty years.
How do you get from exponentially diminishing returns of a dead end technology, to actual AI, in the span of 20 years?

The ai bubble has been growing for longer than all of us know and the big tech companies are now getting the public to fund its rapid expansion through investment and hold some bags in the process.

>>512362390
If you weren't convinced by the clot shot propaganda push that 85% of human beings are already redundant, when they could all lose their jobs and absolutely bring happened to society, I don't know what to tell you.
AI will only replace the flotsam, and let's face it, there's an awful lot of flotsam.

>>512362856
I don't deny that you're redundant. They're absolutely going to kill you. This only ties in with """AI""" in that they're going to use """"AI""" driven drones etc. to get rid of you.

>>512353523
Shut the fuck up, kike.

t. Machine Learning Scientist

>>512353962
For me its Mitsubishi.

>>512362797
>two more decades

>>512353417 (OP)
It dies like every bullshit with the boomers.

>>512353676

https://en.wikipedia.org/wiki/Judea_Pearl
coincidently Danial Pearls father

he created bayesian network and applied them to trading when trading became electronic in the 70s.

>>512353676
If your definition of "AI" includes algorithmic trading, it's a rather broad definition.
Is the printing press "AI"?
>>512353840
Deep Blue defeated Kasparov in 1997.

>>512362725
Programming, data entry, law interns, website 'design', engineering and CAD work. I've read about it, all in the last 3 months, but Fuck you if you think I'm jumping through hopes for a (((poo/kike))).
I work with my hands in the Canadian north, where the air is trying to kill you 7 months out of the year, and anyone that does that, wherever they are, is safe from downsizing. Even jeets are worthless in the field, no matter how many they bring in.
You on the other hand are fucking worthless now, just imagine how worthless you'll be in 5 years.
2030 is coming at you fast.

>>512363063
Bubble sort is AI. Everything they feel like is AI.

>>512363110
Notice how I correctly predicted that you will not provide any kind of objective evidence. You will not do so in your next post, either.

>>512363118
Except it's not AI that is the intelligence, it's the magic smoke.

>>512361198
>it's yugely profitable companies plowing their profits into buying up as much silicon from Nvidia as possible.

That is literally what the AI bubble is about. The world economy puts insane amount of resources and money into AI companies - but the question is whether the AI use cases will go up enough to say that is warranted. Or is there a threshold that won't be passed for a while or worse - will various companies reduce their spending on AI services because of either competition (cheaper "ChatGPT" style AI services) or because they realize that it is cheaper to have many things performed without AI.

The dot com bubble didn't burst because e-commerce or the internet were BS, but because the expectations were too short-term and also too far reaching. The question is, why should AI be different? Will really all lawyers be replaced by AI bots, all HR people, all marketing staff, all teachers, all programmers etc. ... or will AI just stay one of many tools used for productivity increases, thus companies end up scaling back expenses for it if the efficiency gains don't outweight the costs.>>512361198

>>512363110
>I work with my hands in the Canadian north
@grok build a robot that can do this guy’s job

>>512362917
Exactly what is it you do? I can't even tell what country you're pretending to be from you're such a worthless piece of shit, and I'm a fucking leaf.
You know you're replaceable when a fucking leaf has more recognition than even the unknown country you're pretending to be from.

>>512363172
kek

>>512363243
I don't deny that you're redundant. They're absolutely going to kill you. This only ties in with """AI""" in that they're going to use """"AI""" driven drones etc. to get rid of you.

I'm yet to hear any other explanation of how """AI""" ties into the undeniably unfolding scheme to get rid of useless eaters.

>>512353417 (OP)
The bullshit jobs bubble will burst not the AI one. Once companies realize they have dozens of roasties doing busywork that AI does in minutes they will all get fired. You can't repurpose roasties into other jobs so they are getting fired.

>>512362797
>>512362827
>How do you get from exponentially diminishing returns of a dead end technology, to actual AI, in the span of 20 years?

He has it backwords.
Computers were getting more advanced every year.
https://en.wikipedia.org/wiki/Moore's_law

Moores law stagnated around 2016 which led to larger and larger chip sizes.
Inordinate to fill the bubble they first rolled out crypto then ai started to flood the market.

It's no coincidence that these technologies became open to the public right when chip manufacturers started having problems shinking down transistors.

>>512353676
Before the 1987 crash it was called program trading but it was essentially the same thing.

>>512363353
Chuds have demonstrated that a python script could replace roasties a decade ago, but somehow the roasties won.
I'll believe it when I see it.

>>512363189
@grok get that robot to a site in God knows where, and protect that robot from fire, or -40°, or any sort of inclement weather, or a bullet in the middle of fucking nowhere.
Sorry bud, robots are great in the assembly lines/cubicles that you call home, but absolutely worthless in the field, where a huge percentage of all the world's natural resources are, where there's very little infrastructure or people, like where I call home.

>>512363299
Ah, a fucking language model.
See, already replacing 30 jeets and the kike that overseas them.
Brave new world.

>>512363531
Notice how you're forced to keep replying to my posts despite getting owned every time. Shouldn't they at least be handing out shills like you some fake fact sheets in case someone asks you the simplest questions about your narrative?

>>512363412
You are correct.

They don't have the resources to clue everybody in on the mathematics involved so they just blanket the marketing with "AI" and call it a day.

>>512356482
Actually its only limitation is what coding or dev software you’re using. It has a harder time with newer software but if you want to make a game with anything older than 3 years ago, you can do it entirely with an AI.

>>512363451
humans aren’t exactly immune to -40°, fire, or bullets either

>>512362483
Chatgpt 5 and oss. Uo until now, locally hosted llm's have been hobbyists, tinkerers, experimenters territory. We now have enterprise ready tools that can be hosted locally.

We're just seeing the beginning of the AI bubble. Feels like there was a several year delay, this year is the first where I'm seeing visible implementation of AI on a large scale, not just whispers of people using ChatGPT to cheat on tests and give recipe ideas.

It's very reminiscent of early internet, once mass adoption of AI occurs, I think the internet will be this archaic wasteland of old and based on /pol/ being like 70% bots now, I think that will come to fruition within a year

>>512362797
>Computers twenty years ago were primitive compared to the ones now
Not really. The most blatant example is videogames. Crysis is more than 20 years old and it doesn't look that much worse than current games. Sure the resolution was lower (even if 1080p was already widespread back then) but still. Want another example? Look at the weather forecast. Can you say that it has improved a lot in 20 years? Sure the supercomputers are more powerful nowadays but did it lead to better predictions? More power doesn't always equal more efficiency or more practical use cases. Grifting exists.

>>512362856
Pareto's law is absolutely true, but it doesn't mean that the bottom 85% is useless. Life is more complex than that.

>>512363451
True, I'll bow to my AI overlords but only when a robot manages to fix plumbing for cheaper than a human, or can be operational in the field for more than a few hours in the middle of the battlefield. Fuck middle management though.

>>512355892
They make profit per model, but they borrow to research and train the next model. This is because of the intense scale up.

>>512363877
>Chatgpt 5
Show me where I can use and test this fabled ChatGPT 5.

>>512363828
If you don't code your own engine from the ground up in c can you really call it your own game?

>>512363902
The moment 4chun gets swamped by bots they'll pull the plug until they can find a solution. Most likely in the form of subscription.

>>512362765
Not to beat a dead horse too much, but my thesis is that they don't care about inflation; on the contrary, it's not a bug, but a feature, for reasons I stated before.
> We'll see.
.

>>512354107
Son, the inability of a company to turn a profit from a garbage business model doesn't change.

However there were still legal requirements to do with accounting that were still somewhat enforced.
Enron, which famously imploded, ran all kinds of scams including the CFO forming subsidiaries to short their own stock.
Prior to the late 90s, audit firms required all partners sign off on everything and that ensure a decree of proper self policing. A lawsuit changed that and suddenly there was a free-for-all.
In 2001, as everything was really falling apart, September 11 conveniently happened and building 7 collapsed.
Building 7 had housed thousands of case files for accounting fraud.

Fast forward to the 08 crisis. Everything was crashing as Obama came in. So the few accounting rules left that could still keep companies relatively honest were done away with. Namely Mark to Market accounting.

It's been more or less a bubble ever since with the brief exception of genuine panic when government Karens demanded we destroy the global and economy by shutting down as much as possible.

Nothing is going "too fast". Insiders have inflates yet another speculative bubble and they want as many among the general public as possible to be the bag holders for when they sell.
Then all the plebs who thought they were rich are poor again and those plebs will demand the government "do something".
The government will suddenly realize there was something they failed to regulate. Those regulations will further benefit those same insiders and fortify their monopolies.

This has been going on since the 1600s with the Dutch East India company.

>>512353417 (OP)
LLMs fucking suck
They're good at finding information you would have found much more easily if Jews didn't intentionally cripple the Google search engine for 15 years straight

>>512363930
You're being insufferable. It's likely to be released within the coming days or weeks.
You know this tech was an absolute fantasy just 3 years ago? You come across as entitled, untethered from reality, and arrogantly pretentious.

>>512363964
This, but unironically.

>>512353724
They did the same thing to all major search engines
>HEY WOAH WE CAN'T HAVE YOU LOOKING AT EASILY ACCESSIBLE FACTS ABOUT JEWS AND NIGGERS AND GENETICS
I used to be able to find specific comments on Cracked comedy articles by doing a Google search on the post I half remembered. Now that's utterly impossible. Especially if your search query contains "/pol/" in the string

>>512363180
20 years ago people thought the same shit about Sun coming up with shipping container datacenters. "Who the hell needs that?"
Next thing you know, Google, etc. have warehouses stacked full of the things.

>>512364229
So in other words, you can't show me a language model that is to ChatGPT 4 what 4 was to 3.

>>512364229
>>512364333
While we're at it, might as well note you can't show me one that is to ChatGPT 4 or 3 what 3 was to 2. It's a pattern of diminishing returns.

>>512364138
Benjamin Franklin wrote about inflation in his autobiography. In it he basically stated that the monetary system is completely built around using inflationary currency to secure assets like precious metals.

Basically you just threaten people on the terms of paper currency and force them to hand over their precious metals. Its simply means of organizing and structuring stayed violence for the purposes of organized robbery of the population. No wonder he rebelled against the brittish empire.

>>512356482
SORT OF
It can definitely be used to make a game, but the code it draws from is all code written by other humans, and you have to prompt it for what you want and make it all work together
If you ask it for things step by step you can get workable results
If you tell it to create a whole AAA game from scratch it won't work at all

>>512353417 (OP)
nooooooooooooooo, really?
could kikes be lying about intelligence??
nooooooooooooo, never!!!
lolololololololol
kikes scrambling to make it stick.
Hahahahahahahahahha

>>512355820
The mantra repeated on all channels in 1999 was
>it's different this time.

This is a shark telling you to come in for a swim and the water is nice and warm.
People bought Pets.com because they were NPCs who threw money at anything that was hyped enough.
AI isn't world changing either. AI is to distract from the China-tier social credit system the ptb want to roll out everywhere and digital currency so they can control and skim from every transaction. Don't believe the hype.
And Bitcoin is going to be rugpulled so hard it will either go to zero or close to it.
>But Bitcoin isn't...
It's all connected.
Don't buy any dips through 2028.

>>512355698
You sound like a major faggot typing like that bro

>>512355860
>>512355892
It's neither, it's accounting tricks and click fraud that would've landed executives in jail 25 years ago.

>>512356822
I agree with everything except the 100% horizontal growth part. The models improve vertically through larger/more complex datasets, which can only be run through parallelism.
If/when AI ASICs come to be, we will have MoaC (Model on a Chip). That will solve the massive power/scaling requirements. That is years away though. If we make it past the bubble burst, that will be the inevitable result.

>>512363864
But we're a lot cheaper to make, a lot easier to maintain, and they simply can't replace us all at once. They've been trying to use GPS to run the big machinery in the strip mines here for 10 years and haven't succeeded yet, not least of all because of union sabotage.
It's coming, but there's so many avenues of replacement that can be done that are unnoticeable or even gleefully appreciated.
Who's going to mind, at all, if every HR department on the planet gets gutted?
Not until they've got factories for building factories.

>>512364187
>C'est tombe comme le ginseng

>>512356168
Markets and the same handful of companies have been moving in tandem for years. A monkey could Buy The Dip on the Magnificent Seven stocks.

>>512356249
Those keywords that are watched by AI were chosen by tech nerds, not professional writers who by nature understand language much better and have a larger vocabulary.
Even "earnings" has myriad variations. I actually worked in this specific area for a media company's own internal product and so I know the real limitations probably better than anyone.

>>512363433
This. They dont pay them for their labor.

>>512364767
True true

>>512353523
It’s cost/benefit. They aren’t selling corporate subscriptions that justify investment in hardware for the big data centers. Nobody needs everyone to write their emails 2x faster enough to pay what is being asked. It makes communication worse, you have two people using chatbots to communicate with one another and neither human knows what the fuck is going on. Nobody is willing to pay for the funny meme image generation. They haven’t found a way to monetize it. Startups are promising the moon. Two more weeks and we’ll have self driving cars and truckers will be doomed. Two more weeks and we’ll have combat drones that don’t require human intervention. Two more weeks and we’ll eliminate creatives in art.
What it can do really well is organize big data sets and summarize them. It fucking sucks at human reasoning, but it’s good at math. If you ask if what happened in a story it tends to not know. It fills out a form with words that repeat a lot. It can kind of pick out themes. It’s just terrible at understanding emotional concepts.
So just like the dot com burst you will see the chaff separated from the wheat. There will be companies that stand strong. Yeah, stock trading, quick data set summarization at colleges. More user friendly versions of shit that search engines and stat engines can already do. Is Microsoft going to keep pushing CoPilot with every license of their cloud-based office (that is floundering by itself because it doesn’t work and big companies are rolling back to on-site)? No, probably not. Not unless hardware costs on that go significantly down.

>>512364312
This. Nowadays you generally see Reddit or Quora links on top of your search page when you're simply asking a question. Those sites are also the main sources for LLMs, except for the "official" news sites and Wikipedia. The rest of the internet is getting invisibilized (the word probably doesn't exist but whatever). It's hard to stumble on a small forum link on the first page unless it's a very niche subject.

>>512364573
Internet was world changing, AI is also a world changing tech, but it will take time before it delivers all its promises.
>Don't buy any dips through 2028.
That I agree, congrats to all those who bought at 16k or below and didn't sell, but now it's time to take your profits if it hasn't been the case yet, unless you're a maximalist who plans to die with all his BTC in a ledger next to him.

>>512364744

You probably know more about it then I do, from what I've seen AI models aren't developing more complex cognitive processes necessary for higher levels of intelligence but I may be wrong

I mean I'm sure it's coming but from what I understand most AI companies are looking to scale up with their capabilities of pattern recognition, not to develop meta-cognition or anything actually approaching human intelligence

I can see that happening if we manage to optimize hardware side of AI computing like we did with computers in general and therefore be able to start congregating basic cognitive processes into something truly more advanced

>>512365099
>Two more weeks and we’ll eliminate creatives in art.
If you haven't noticed the attached field is in absolute ruins right now.

>>512353523
King of fighters 2000

>>512353768
retard. ask a help desk chat bot what the date was three days ago.

>>512364396
I didn't know that. It makes sense, bad money drives out good money.
Gold is VAT free in (all? of) EU, at least buying it is. I'm still not sure why is it still untaxed, but it would be fun to see it taxed in the near future.

>>512363433
You can't use a simple script to sign stupid paperwork. You need a human to fill it, sign it and another human from a third party to verify it. I work at a small constuction company and every time we go to a new cnstruction site to work we need to fill dozens of documents for each worker. Most workers don't bother signing them, the roasties at the office just falsify their signatures and upload the documents to the third party website. It's way quicker this way.

With AI you could upload all the documents necessary with their own signatures and they would get revised by an AI and give the ok. No humans involved. My fear is that this increase in productivity would get gobbled up by the government in one way or the other. Just like the internet didn't slow down the growth of government bureaucrats despite the increase in productivity, we might experience another increase in bureaucracy by the government. Now they are going to ask you for certificates to buy bread or other useless shit to justify their own existence.

>>512356948
There probably will be another crash because global finance is so interconnected that any real break in one area will spread like wildfire, or like an arc along a power line.
I'm betting on or being Bitcoin because if Bitcoin collapses there's a case that all investor losses are deserved. That it was the domain of drug dealers or tax evaders or that nothing really backs it, etc.
But the so many companies and pension funds and governments are now loading up on it.

My guess: the rulers want another system already in place.
They make as many people broke from all these schemes and just as they're about to riot, this magical savior of UBI and digital currency comes along.

>>512365266
Making free meme images on someone else’s hardware isn’t sustainably monetized. The access is only there to generate hype.

>>512365367
https://en.wikipedia.org/wiki/Exchequer_of_Pleas

https://en.wikipedia.org/wiki/Star_Chamber

https://en.wikipedia.org/wiki/John_Law%27s_Company

>>512357379
How many of those layoffs were people doing actual work vs lazy girl jobs?
I've read about HR being replaced but that entire profession was worthless anyway.

>>512365099
I mostly agree with your post, but
>It fucking sucks at human reasoning
You could say the same for a lot of humans. Most people are dumb as fuck, and need time to learn new tricks and processes. In an environment where it's becoming harder to recruit capable people thanks to the failed education system, AI could be a nice band-aid for a lot of companies.

>>512365367
VAT and other taxes come when you sell your gold (=take your profits). This is why it's not taxed when you buy it, it's considered as an investment. It's exactly the same for crypto.

>>512353523
>can’t remember things
>can’t learn things
>can’t reason
>can repeat retarded pajeet quora posts

In a sane society AI would just be deported on the same ship that takes away every Jew h1b and black. Happy even to make it a luxury cruise deportation if that makes it more palatable.

>>512354131
When skill is "free" what is to stop anyone from just DIY and cutting you out of the loop?

Monopoly? Intellectual property? GOOD FUCKING LUCK NIGGER,

You tards are 85 IQ in a 60 IQ world.

>>512353417 (OP)
Ai has never overcome the induction test, logic breaks down into deduction and induction. Computers passed human capacity for deduction basically at their conception, they are getting better and better at it sure. But they still can not induct knowledge, and induction is not only the root of all knowledge but therefore the seed of new knowledge.

>>512353523
>Improving AI
>Runs well on laptop
Runs what? Itself? So it can guve me funni answer for free?
WHAT DOES IT DO that would justify me spending shit on it? Why can’t you give one non-retard example instead of saying “it works well”?

>>512365618
Youre not wrong but the things people are generating on their own hardware isn't much different.

Which itself isn't much different from what people were making with photoshop. Hell even photoshop has AI components built in.

>>512358593
>jews
The iron curtain came down for one generation after WWII.
Long enough for Jews who survived Auschwitz to not being able to reunite with any family on the other side who may have also survived but told their family was dead.

>>512358707
Railroad bonds were a bubble that destroyed a lot of people's wealth in the 1800s.

>>512365516
I think you're right, or at least your scenario is viable.
BTC is the easiest vector - drive people from fiat to crypto, then embed some CP in ledger, dump all yet unused IPO BTC and issue official crypto with initial conversion rate, say, 1:50.
UBI will be the great pacifier, especially if it will be introduced in most White regions first for everyone else to see and to envy.

>>512365704
> John Law
Thank you, I was searching for exactly this.

>>512365722 checked
Not really - no taxes if you keep it longer than a year (can't remember exactly). Besides, silver isn't tax free in EU, except in Latvia and one other country which I can't remember either.
t. bought gold with crypto maybe 10 years ago

>>512353676
Algorithmic trading has been around since the 1980s or possibly earlier.
Machine learning probably got incorporated in the 90s.

https://en.wikipedia.org/wiki/Program_trading

>>512356435
lool, there is no way with current hardware we could get anywhere NEAR the level of pattern recognition needed even to catch and throw a ball, a simple task for us but until you can achieve it with AI and robotics in a repeated open environment, we are safe
People have zero clue about this shit

>>512366614
>People have zero clue about this shit
Indeed.
https://www.youtube.com/watch?v=htXUNnrUFOs

>>512364138
Aristotle wrote that tyrants deliberately cause inflation because people struggling to afford basic living costs don't organize against them.

>>512364201
Unironically a skill issue. You probably use chatgpt free version and ask it to make loli porn and get mad because it won't, then come here and reee about AI being dead. 90% of AI doomers are like this
>>512364377
I have a feeling I could show you everything and you'd still deny it. Just admit you're trolling, it's pretty insufferable at this point. The data is there. I'm sorry you're getting stalled because you can't break out of the chat feature, but for real world applications, AI is getting better. Go back to writing sonic loli fan fiction in your basement and leave AI to people who will actually utilize it properly.

>>512365220
The world changing aspect is block chain, not AI.

>>512356822
Good post, anon.

>>512365704
Interesting stuff.
I was reading into gold manipulation in the 1920s recently.
https://www.lbma.org.uk/alchemist/issue-93/what-happened-in-the-london-gold-market-during-the-gold-standard-1925-1931
https://www.lbma.org.uk/alchemist/issue-92/the-formerly-missing-gold-price-data-1919-1925

>>512366361
A lot of people blamed the 1987 crash on it too.

>>512367170
if you think this is even ANYWHERE near what I am talking about, get into the clueless line

They are throwing the ball in a complete sensor environment, the ball is tracked within that environment and the robot makes a catch, millions of pounds to achieve in a completely sterile environment that is impossible to even think of recreating outside.
its a party trick, with a ball, pre-programmed route and velcro hands so the ball sticks, not AI

>>512353523

> please use shizo loopback machine

lol

The biggest problem with AI is its application
Literally, what do we use it for, to talk to, thats pretty much it
Will it immediately take over jobs, some but not as many as people think. In order for companies to use software, especially AI it has to be insured. You want to try and convince a board of accountants you have software that will save millions, job done before you finish the sentence, try it with the lawyers, its over before it begun.
The odd power blackouts etc, the supply chain errors, transport/air controller problems
I can tell you that AI has something to do with it and its also the cause of what will most likely turn out to be the start of an 18 month stutter in development, people are seeing errors and are getting scared (as they should) but its errors in code and application of AI, not AI itself
when you think of the implications, it should be more terrifying than AI itself, imagine some code monkey in his data center that he is admin off, he has his isolated sector of hardware (only the best, its not his money, all he is doing is using his "test" environment) and running an AI, he is looking to make a quick buck, so is looking to develop(not use, just for sale)hacks, be it hardware software etc he is just playing, the AI is running through iterations and it comes up with a variation on a legacy hack for older hardware, he wants to test it but because he fucks it up, the AI starts replicating and targeting the data centres networks that lead back to client facing networks, you get the idea, thats how easy it is to fuck up with AI.

>>512367529
>I have a feeling I could show you everything
Go ahead. You literally won't. I know your entire dialogue tree and every line of your shill programming.

Your next routine is a 50/50 chance decision between deflecting or ceasing contact.

>>512365365
last time I asked bing chat she didn't know either

>>512353523
>There has never been a tool in human history that has ever even come close
to having this much time, money, and effort poured into it for zero beneficial results.

>>512353417 (OP)
It won’t.
And stop calling it AI, it’s machine learning. That shit has been around for a while.
Auto correct, for example.
ML is just getting started. Stop focusing on LLM or GenAI and focus on things like image processing and other models to help process large quantity of data.
It’s not a bubble, it’s the electricity of the future. Large part of that ties to having the proper hardware to parallelize those calculation aka GPUs with architecture designed to do it and do it well.
Stop being a Luddite.

>>512353417 (OP)
its not a bubble but productivity multitplier

dotcom was a real bubble because investors pay millions for shitty sites, today I make job done thanks to llm and have personal llm teacher for some things

you need to have fuckin power plant for making models and AI hardware which one cost tens of thousands USD, at dotcom time every faggot with HTML editor was a "startup"

>>512369108
*replicating
To clarify this point, I am talking about the AI repeating a simple query to all the hardware of the legacy type it can find both in the data center network and beyond, just to see if it would work, this simple fuck up can shut down entire networks just because the network AI protection sees a potential intrusion. just one of the ways AI can fuck your day up

>>512365099
>>512356822
Agree with both of these. The bubble is based on the economic model supporting it. At best, we are early in the enshitification cycle where they are heavily subsidizing the technology to drive broad adoption among the consumer class. Next, will be a value transfer to the businesses at the expense of the consumers to drive business adoption (all that data they are harvesting to train the models, plus all the data users are willingly sharing with the models become the business commodity). Then they need to gouge the businesses to take as much value back for themselves as they can before the platform collapses. And that's best case. That doesn't take into account any potential aggravating factors such as model decay, prohibitive energy/infrastructure costs, etc.

>>512353724
noticed this with enterprise/volume licensed claude and grok. chatgpt has improves. I dont discuss anything non-technical with AI (and triple check everything, use it more to deep search and suggest relevant papers and obscure docus as a "rubber duck assistant"), so even though chatgpt is woke or whatever, that doesnt affect my work.

>>512358635
https://en.wikipedia.org/wiki/List_of_unsolved_problems_in_mathematics

Orly

2026

>>512369622
This anon gets it.
It’s not the silly chatbot at the moment that is impactful. To run any models at production scale requires tremendous hardware and power grid backing.
Most people are unaware of the land grab that happened silently several years ago.
Look at who purchased the most nvidia chips: meta, Microsoft, Google, Tesla, Coreweave
It forced chinks to come up with memeseek but that is only based off of the pretraining OpenAI already accomplished.
The goal is to do it like AWS where you can loan a small fraction of the processing power necessary for your business because you’re too fucking poor to buy a bunch of them or have the power to enable them.

>>512370125
its the restriction of access to live data, thats whats causing most of it, mainly because the devs are shitting themselves about litigation, from wide and varied areas of industry, especially those that involve manipulation of live data

>>512369622
except this time productivity isnt resulting in greater access to durable goods. food, housing, and medical cost MORE.

the "productivity" is tail wagging the dog garbage of more AI generated artwork and thousands of calendar apps reinventing the wheel while requiring exponentially more processing power and energy.

you are defining "productivity" as worthless karen sinecures being able to write more AI emails to be read by AIs on the machines of other karens. it's all bullshit, trying to spawn "productivity" out of nothing.

>>512370323
Nearly all """AI""" enthusiasts are genetic dead ends who, up until 5 years ago when the """AI""" psyop got rolled out in full force, were planning to rope. You are trying to rob this person's only hope for the future. You are trying to take away from him his AI god who promises eternal life in virtual heaven. The AI god who will genetically modify him into a chad, make him immortal, give him full-dive VR into fantasy kingdom, give him a harem of AI-driven 12 years olds, give him the 60 IQ points of intellect that he lacks, etc.

These people can't be reasoned with because reason wipes out their happy future and lands them back in the suicide-fuel reality of their lives.

>>512353417 (OP)
Doesn't the AI bubble pop every few months when the chinks leak an ooen source version of the current models

>>512355820
>a computer can do math
Wow, I'm truly shocked.

>>512370730
irony being that it is the entire point of using it. for example, proquest's AI paper search and analysis for searching the corpus of human knowledge, or westlaw or cocounsel for legal.

non-logical LLMs are great search engines.

>>512370994
exactly
I have had to include, by default, a function in my searches to enforce a check of the most up to date information, this has changed in 3 months. I never had to, then sometimes, now its utter bullshit, I have had arguments with a search engine because the data is transparently wrong, just down to not searching for up to date data, a restriction that the AI in question admitted had been increased.

Most people complaining about AI having problems don't even understand how versatile machine learning is and think it's all LLMs and video generation. There's models that are trained in weather data and can run global forecasts with 1x1km resolution on phones. Nobody talks about that stuff because to them it's not actually 'real' AI but it's the same technology that you see in LLMs, just applied to different data. It's here to stay even if Chatbots never replace software developers or video generators never replace animators.

>>512370865
That's one hell of a depressing projection anon

>>512371591
Correct.

>>512371269
yea. what I do is simul chatgpt (best general engine + claude or proquest or something), paid model and with search and deep thinking or reason mode, with directive to cite itself inline. then feed it across models. so they "check" themselves and each other.

sucks that LLM+logical hybrids arent popular. it's becauswe our faggot overlords want to usher in self-generated AGI which will be a nightmare.

I hate all of this.

partly my fault though, I was an AI researcher during my phd a decade ago before transformers got popular. CNNs, RNNs, GANs.

>>512371591
the worst aspect of AI aside from environmental is that midwits think it is a God, and cant evaluate it as a tool, and cannot evaluate when others are similarly misguided. blind leading the blind generating all garbage with no checks. fake people doing fake work for fake money and buying fake products. it's like airdropping weapons into subsaharan african villages. people in general cannot wield this power effectively.

>>512371622
It's a well-grounded observation. Whenever you see these types get together sooner or later it devolves into wankery about AI wives, extreme escapism and schadenfreude about what they imagine will happen to women, talented humans etc. when AI god makes humans obsolete etc.

Who are you trying to fool here? It's the ideology of heavily dysgenic spiteful mutants.

>>512353523
what a time to be alive~

The type of AI that would transform the world will require so much energy and fresh water, it will not be feasible to implement until a breakthrough energy source like fusion is available.
At least 10 years off.
Nividia doesn’t want you to be aware of that.

>>512369217
K

>>512353676
Statistical models aren’t AI and neither are LLMs

>>512372124
>Usable AI is 50 years in the future
3 years later
>World changing AI is 10 years in the future!
See where this is going?

>>512353523
Don't listen to this guy. He doesn't even have his timelines right. OpenAI released those models yesterday, and it's unlikely they'll outperform existing open source models.

>>512353417 (OP)

Honestly I hope the world economy does collapse. I wont need to go to work. I can grow food and hunt animals and cook them on an open fire, and any lurkers who approached can be removed without consideration to law.

>>512371769
not as far ahead of you but my own way will be to use a set of the different llm's and look to have an AI search API(unfortunately, whatever one is least restrictive, which will obviously have an effect) that they can call rather than just the library data, this will go into a central index, along with chats and sources gained from use.
Don't happen to have a spare £200000 do you?

>>512353724
>guy in the middle doing hover hands on the show of only gigachads and megastacies

>>512353417 (OP)
One can only hope

>>512353523
Found the retard who invested all in on startups

>>512353523
LLMs will never be AGI, anon. odds are, AGI is impossible and may even require quantum computing to work (which is decades away, if ever).

the tech industry is not being truthful about "AI" for obvious reasons

>>512372276
haha I wish I had extra cheddah. I do think there are those AI toolkits that run simul queries against each other. I dont use AI much though or follow it too much aside from oh model update. which is why I just copy paste between them when I want them to look for something. meanwhile I also will be thinking and using a search engine

>>512372333
>Found the retard who invested all in on startups
Found the retard who thinks retail investors can get in on startups.

>>512353962

I think it is impossible to remove hallucinations from a word based LLM model because deep down they are recursive text generators running off token probability in context. A LLM can talk about a topic but it has no real concept of what the topic actually is so it cannot filter out hallucinations. They are essentially bullshit generators which is why they are so impressive.

>>512372493
its still shit in general at the moment
a lot of it is due to the step backs in live search
and also in the cutbacks in learning models, people dont realize how much money this costs and its literally dead computing time

As I said earlier, we are probably looking at an 18 month slowdown, I am trying to get myself positioned for that

>>512353417 (OP)
>if it pops it could take the entire economy with it
Then it won’t pop lmao. They won’t let it if it’s going to tank the entire fucking economy

>>512353962
>>512372717
I have noticed a drop off in quality as well and I put it down to one thing mainly and its restricting live data access and trying to use its learning data way more than it should because the devs are afraid of litigation

>>512371956
>It's the ideology of heavily dysgenic spiteful mutants.
It's not an "ideology". Technological advancement is imminent, and it's merely an observation. Your ad hom is out of place, Edward.

>>512353523
I don't notice anything magical or turbo improved. Maybe it runs a little cleaner and a tad more lucid but that's about it. Hardly worth the hype being thrown around the market. It's fine, I like using it, but if it disappeared I wouldn't miss it. It's mostly a toy.

There should be a quiz before anyone posts about AI. If you're only experience is chat interface, you can't identify a use case beyond asking it consumer tier requests, and especially if you don't pay for it, your post should be immediately deleted.

>>512372872
Lol case in point, this retard probably hammers away vague, limited context questions at Gemini free and gets back garbage, then complains its all vaporware.

>It's a fucking bubble!
Seethed the desk jocky.
AI has made so much progress in just the last 2 years, it's clear AI is the next revolutionary tech. We are talking Horseless Carriage levels of importance.

>>512372856
>It's not an "ideology"
>immediately spouts the dogma of his ideology

>>512356690
>2. Lowering model precision (FP16 to FP8/FP4) to cut costs

Lowering the precision reduces memory footprint and cpu computing consumption enabling larger models. Its not just about cutting costs. Hallucinations are unavoidable because AIs are language models.

>>512372726
well said. to add, people also cant fathom how much energy is burned on useless shit. rolling blackouts in some states? needing to frack for oil and NG? what a waste. nukes, hydro, geothermal, and eliminate half our datacenters.

>>512373004
no you're the midwit who thinks that your boilerplate app or ai artwork is revolutionary

LLMs have their uses (contextual polysemy) but they havent improved much, and some have backslid somewhat even while requiring exponentially more compute (at least for training)

>>512362661
MFW I do DevOps and AI can't even have a crack at 99.9% of my tasks =D

Yet I can read daily in the news and on imageboards how virtually all programmers will be replaced VERY SOON =( I'm sure all the women journalists and imageboard basement dwellers are right, and I am very fearful for my job!!!

>AI bubble
Literally a bunch of retards jumping on shot tech. But maybe people felt the same about the internet bubble in the past too. This just makes me think of that image where the guy puts a stick in his bike's wheel while riding it.

>>512363110
>engineering
:D

Wanna take a guess on what percentage of your car's engine was designed by AI? Or the drill you use at work. How about the magical card deck sized device you have in your pocket that transmits data 1km away at speeds of 1gbps?

Oh right, you have no idea how any of those even work. But hey thanks for doing all that manual laboir for us. I couldn't be arsed personally :D

>>512373576
>MFW I do DevOps and AI can't even have a crack at 99.9% of my tasks
Have you even tried?
Claude is really fucking good.

>>512373141
>thinks an observation is an ideology
Smartest Moldovan

>>512374265
Can Claude do C#? I use Deepseek and GPT currently but it fucks up once scripts become too complex (over 1000 lines). Would you say its better

>>512373809
Cars currently arent really designed by AI but still mostly traditional methods like Cinema4D Blender CAD or other likewise software, the drill bits are literally designs from decades or hell centuries ago fact of the matter is you obviously dont know how any of that works which is why you unironically believe AI designs the majority of car engines or that this is some industry standard peak boomer midwit

>>512374466
>Can Claude do C#?
Never tried, but I'm sure it can.
>I write scripts with over 1000 lines of complexity
You honestly sound kind of like a shit programmer.
You should be writing small functions and classes with well-defined interfaces to build up larger logic.
Read up on the SOLID principles, ffs.

>>512374766
Youre correct generally i dont make them complex but occasionally for example for a voxel game it can get pretty complex same for base ai systems, aswell as large worldloading and saving systems, besides that my code rarely exceeds 100-200 lines but its exactly these few complex things in which AI becomes unreliable anything below 400 it does really well though

>>512374466
I hope you meant 1000 lines of all your libraries and implementation added together.
If you’re a megascript enjoyer then perhaps you’re due for replacement.

>>512374931
Well for example i have a script for my voxelgame that allows the player to switch gravity modes which then aligns the players local space down axis to a new direction when in gravity mode this also handles movement jumping crouching ledge behaviour, whereas things like Health are their own seperate shorter simpler scripts

>>512353417 (OP)
It will be like the .com bubble.
As long as you're invested in the industry leaders you're going to be okay. If you invested in some street shitting Indian scam project then you're going to lose your shorts.

>>512375053
Sounds like an architecture problem.
Never done gayme programming, but it seems like there should be a better way to accomplish all of this without needing to resort to large / complex logic.
Sounds like a fun project, in any case.

>>512375053
I'm not trying to be mean, but you should train and use the mush between your ears for this. LLMs cannot synthesize information in a logical, determinate manner

>>512375496
O I do, those large scripts i write myself into working conditions, ai stil nice to lay a basis but ive been coding in C# long before AI

>>512375481
Well its mostly because its a giant procedural survival game with allot of content and systems working together aswell as physics calculations

>>512375481
For example I have an inventory system that allows for crafting, 1x1 blocks and larger s.t.a.l.k.er dayz type 3x2 or other shapes (for rifles and such) ontop of this additional code for dropping items outside of it or to delete it etc. That does kind of add up but its completely functional

>>512375592
>Well its mostly because its a giant procedural survival game with allot of content and systems working together aswell as physics calculations
Are you using a game engine?
I hear Godot has gotten really good.

>>512375546
good, yea boilerplating and asking for refinement suggestions is where AI shines.

good luck with your game