An M.I.T. study found that 95% of companies that had invested in A.I. tools were seeing zero return. It jibes with the emerging idea that generative A.I., “in its current incarnation, simply isn’t all it’s been cracked up to be,” johncassidysays.bsky.social writes.
I suspected.
📌
LOVE TO SEE IT
My son decided that our family has to boycott companies who apply AI to replace human beings. I haven't seen a new movie since.
You're describing my friend and her son... thanks for the laugh... you've earned a follow
AI was treated as a joke for a long time, but is now over-hyped and showing signs reminiscent of the dot-com boom a generation ago. We quickly learned that just having a website was irrelevant if there were no rational business plan. Similarly, many 'AI-powered' enterprises will fail spectacularly.
You need to be using it not it using you Like phones and credit cards and politicans
"IA" is a huge fucking waste of power.
Don't tell The Market that, they don't like to be corrected!
Pump and dump baby! You wanna buy some NFTs?
AI reminds me of backseat/frontseat yapper 'helpers' when I'm driving. Just zip yr lip while I'm watching for nutbars tailgating/careening fr tiny openings one lane to another so they can get to their destination 30 seconds faster. Unless I initiate AI/navigator for help, stay outta my ear/eyeballs.
Obviously.....The 'A1 Investment' is Nothing like the Tech BUBBLE..... Nothing.
Nope. And it’s sucking a ton of our power.
Yep. Wasting electricity, water, poisoning the air, ruining communities, making people sick, raising our utility prices....
Surprising someone, somewhere, but not here.
👏👏👏👏👏👏👏‼️
It's ok at drafting templates. About the only use I've found for it
The maturity of AI right now is about equal to a petulant two-year-old. A lot of what is being touted as AI is actually little more than Excel.
Another soup-prise!
Searching just sucks with AI now. I’m 3 pages into the search now before I find an actual expert
It can't even give me the right phone number for my doctors office
Because Americans just love being forced to talk to and engage with “canned” thought for every aspect of their daily lives. Try calling any company you do business with or using their online computer “assistant”. Exercises in frustration, aggravation and rage. SISO… stupid in and stupid out.
No problem. They’ll just steal the money from us to break even.
hiddenmatter18.blogspot.com?m=0
No shit morons. Its MILITARY ENGINEERING It's not designed to generate PROFITS.
AI requires a lot of time and effort to train up before it'll ever be useful for anything. Unfortunately, no investors will ever be happy funding something that takes time and effort while also costing then a lot of money to develop fully before trying to make a profit from it.
My platform has gone all in pushing agentic AI to customers. The problem is that the small businesses that would jump in can't afford it as it's priced, and the large enterprises are risk adverse. That doesn't even take into account businesses in regulated industries that can't adopt.
#noshitsherlock
I have only found AI to be a waste of time. I absolutely hate communicating with AI on the phone. The only good thing to come from AI, is gps.
Artificial Intelligence is actually remarkably stupid based on what I’ve seen from it.
There is no AI: what we have are programs that do as they are told; interpret data as they are told, ... That Is not Intelligence.. Intelligence is the ability to make RATIONAL decisions based upon what your past has taught you.. We are flawed, most people are not Rational: AI is a hoax
current A.I. is nothing more than an updated Search Engine.. and not even an improvement - it contradicts itself based on how a question is poised to it. accurate information is not the end goal.. increasing interest, and therefore "sales" is the current map.
So true. This "AI" is nothing but a frustrating mirage of what people promise it would be. We wanted the Doctor but all we got were inferior Daleks.
It might go one step beyond and produce something concrete on your behalf - possibly something of no value or downright incorrect. And how does that add to your knowledge or skill? We're turning from couch potatoes into potato heads.
a short trip for some, lol
Seriously. You can't get anything useful from Google anymore since it changes its answers each time you rephrase the question or ask it to expand on the information it just gave you.
Been thinking this for some time now. Looks like we've (they've) hit a wall of diminishing returns. Some speculation that the current approach of just throwing more compute power at it will only yield small incremental improvements.
I'm actually glad to hear this.
Zero return and turning a blind eye to the environmental harm caused by AI.
I saw a chart of ChatGPT usage where the usage dropped off a cliff after June. It seems the primary use case is cheating students. Big win. Let's butn down another rain forest so Timmy can get a C+ on his geography quiz.
The only other use I’ve encountered is AI generated porn. When you consider how much porn AI models have scraped, it would seem AI porn “artists” can churn out a few thousand images and flood nsfw websites with it daily.
Final marks need to be sit down exams... if you don't understand the content, you aren't qualified. Any 🤡 can ask AI for crap answers.
So correct. I don’t understand why we let these selfish bastards / tech guys destroy the earth for their game.
Yet, so many students were caught, lol. Dumb is as dumb does.
No fucking shit
Most chatbots suck, but AI driving cars, landing planes🤪 and teaching our kids is ok.
innovativeblogtech.com/airtel-perpl...
It’s pretty gross that AI is essentially serving up a pseudo-ad like this.
Zero return, or actual losses in productivity, profit, etc?
It's probably this, to be honest.
It’s great if you like people with six fingers
I just see it being abused to promote a ton of crap. People see it an just switch it off.
And yet it's going to suck up a lot of our energy and water resources (or worse ratchet up nuclear power) all for mostly meme generation, consumer data harvesting, and bitcoin mining ... 🤔😬
Here's my guess at the commerc model: release free use systems that will collect huge datasets from their users Slowly enshittify free use systems to increase affordability and continue data collection Move to expensive commercial accounts for unshittified AI products improved with end-user data
"Longer than expected" like... NEVER 😆😆😆
📌
Computers revolutionized our world. AI will do the same.
Yeah it's cool the program can generate answers but really the most effective uses of LLMs I've seen, which predate the current hype cycle, are: 1. Look at this image and tell me if there is an abnormality compared to other similar images. 2. Look at this code and tell me if there's errors.
By nature of using stochastic models, generating answers to questions or images are by their very nature random and only have a reasonable probability of being correct. Until someone figures out how to combine deterministic "true" guardrails to these models, any results generated are suspect.
Yes but they are correct more often than not. I use them for all sorts of things. Financial research, medical research, website rewrites, etc. I check the answers as I’m aware they can be wrong. Of course they are still in their “infancy” but as time goes by their accuracy will improve.
Oh for sure they will improve over time. Unfortunately though, in my use cases, I've had better results with deterministic machine learning tools. We've tried some stochastic models but it was too risky (healthcare bioinformatics). The LLM was maybe 60% on target, the deterministic model was 90%
the improvements also seem to be *crazy* expensive - how much has been spent to get models to where they are now, where anything important made, you'll want an expert to check anyway, often for minimal or no time-savings? How long until it actually makes a profit?
Tbh I don't think this iteration of LLMs will ever turn a profit. Capital is desperate to create new markets to exploit and this is just the latest in a long line of weird attempts. The bubble will pop soon and with luck red hat or whoever will make a useable FOSS LLM through normal software dev.
yup - there's some value to them, with better auto-complete, improved typing etc., but there's are marginal increases, not a multi-multi-billion dollar service, bigger than smartphones! Especially as they need super-expensive upgrades to improve, rather than being one-and-done software
LLM’s are like the two hunters running from the bear. One says to the other, ‘I don’t have to run away from the bear. I just have to run faster than you’. LLM’s don’t have to be perfect. Just better than humans
The subscription ones are very accurate. Not a huge shock there is it. It’s just software for now but one day … They are definitely playing with fire. lol
They are excellent for financial research. Having it spit out 25 year projection models feels like magic. It beats me doing all of the math in Excel.
Exactly. And they are great at sniffing out debt loads and interest payments.
Be interesting to see how they stack up in 25 years. LLMs are terrible at math.
Even a broken clock is right twice a day
It's a bit like why torture is useless for intelligence gathering. No matter what answer you get, you still need to confirm it with actual, old fashioned shoe-leather intelligence gathering.
While many general purpose AI systems are generally trained their memories contain falsehoods from the training material. Hence reducing trust in AI.
But a bunch of idiots on social media who have no idea how to use AI or any of this works Tell me it's the devil. Are you saying they have no idea what they're talking about? Why would anybody with no idea what the fuck they're talking about say so much over and over again?
That depends on whether you count "making the next generation of people incapable of running civilization" as a revolution. Because that's the main observed effect of LLMs, so far.
How did you feel about computers?! I mean basically you have a point but AI is here to stay. they will put lots of people out of work just as computers did. We need those jobs. However there is huge investment opps with AI
Look back up at that 95% figure. LLMs *do not* make money. They are, demonstrably, not useful. The spinning jenny at least produced a lot of thread, the Apple ][ could run a spreadsheet. ChatGPT just drives tech executives mad and lights hundreds of billions of dollars on fire.
I don’t know who came up with that 95% number unless you are referring to some paper written by MIT specialists, I think it was? However these guys are notorious for getting things wrong. And it begs the question just how could they have arrived at that percentage in the first place.
Source?
Really enjoying the quality of my students absolutely plummetting through the floor since their lazy asses obtained access to the "do the work and, more importantly, the thinking for me" machine.
Hi, I'm Doctor McChatgpt, what seems to be your trouble?
No. It won't
Yes it will, investments in infrastructure, data centers and development of better ai is estimated to run 3-4 TRILLION dollars.
😹😹😹
Maybe AI will, but LLMs aren’t AI. There’s some applications for them, but people are very quickly finding out they’re not actually useful in most applications, so the continued shoving of them into everything has been a colossal waste of resources and time.
I don’t know. Maybe don’t use them till the tech is made better. To each their own. I do find the subscription versions to be more reliable. (no huge surprise there)
LLMs are a dead branch. They will never be useful for everything, only very specific services. But because that’s the path that VCs have sold for untold billions, that’s what’s getting shoved into everything even though it clearly doesn’t work.
WOW !!! Imagine that….
and worse
Totally no surprise. AI is no magic bullet.
Firebomb their data centers anyway!
In AI right now most people are basically trying to go to the moon with stone age tools. But their failure does not mean, that there is no future for these tools, just not jet for that purpose in the next few years. Simply use AI tools within their current limitations. Propably they will evolve.
Thanks. Rarely something to read between the extremes.
I've had to use it and I'm not impressed.
BECAUSE NO ONE. EVER. WANTED IT.
That’s a lot of words to say: it’s a freaking bubble.
AI, tools for lazy people.
Of course it’s not! We’ve known this for a long time - it’s the tech sector and media who jumped on the bandwagon and never asked questions.
AI sucks and I hope it dies an ignominious death.
A long time ago word processor systems in business were supposed to reduce the amount of printed paper. In fact they increased it. Because authors would edit the printed version then ask for an updated document - repeatedly.
There are people (entire offices as well) who print out their email … 🙄
"BRO! there's this new thing called the "intronet" an no one really understands why or how it happens but alls ya gotta do is get yourself one of them dot coms and sell something and because it's new and poorly understood everybody will throw money at you and we'll all get rich!" Same song
Has your company prepared for Y2K yet?
Dude, you know it's going to kill the computer in everyone's car. For real!
AI s a bubble. AI is becoming a commodity. After DeepSeek, there is no way to keep the technology proprietary, and so the price of AI services will be driven down to cost. No margin. Which means tech giants have misallocated billions. www.jandehn.com/post/the-ai-...
📌
Another AMERICAN tech bullshit sales job.
Only thing so-called "AI" can do is process information in an automated way. I have always preferred, therefore, to call it "AIP" or "Automated Information Processing". If you need large volumes of info processed in an automated way that would grind humans up, here it is. But it's not "intelligence"
Good. Jumping on the latest tech guarantees nothing. AI is interesting, but I've grown tired
Same as self driving cars? Not long ago ubiquity was just around the corner. Then it ran into a wall of real world reality.
So much hype and what does it give us? Fake videos, fake photos and sod all else…
SYNERGIES!
If only they had read & took Asimov in mind.
Help me out!
I am not amazed by how little Ai generated. It's worse at predictive text than before it was invented. It's worse at searches too, the first page is no longer a guarantee of finding what you wanted; unless you wanted things to buy. Ai is a shorter word for scammers with Ai voices Ai faces.
The market demand for confirmation that Gen AI is useless, expensive, and doomed to failure is strong. It isn't true, but people want to hear that, and the market is serving that need. Read the actual MIT study for a different story.
Yup! It’s also part of an odd kind of vampirism, seriously! 😳 open.substack.com/pub/noelkeit...
Gosh, who knew? (A lot of people.)
Is it the j curve that all revolutionary products experience?
Screw AI!
Pro tip; never invest in anything you don’t understand. AI, Bitcoin come to mind.
That’s an excellent tip. Here’s another; study bitcoin until you understand why it has value.
bribing officials? Intentionally crashing the us economy so you can buy it for pennies and enslave the poors while genociding poc, disabled and lgbtq? Bitcoin was ok for drugs on silk road. bitcoin is the currency of choice for shitty people
Your world is filled with enemies, boogeymen, there’s evil all around you. You and your comrades are warriors, fighting the good fight, it’s become your identity, a hero. Your information comes from like-minded men and women, just like you. There’s no hope of you ever seeing bitcoin objectively.
fyi bitcoin users think it cannot be tracked, but it absolutely can. hope it is worth crashing the economy
So why do you folks always imagine all these bitcoin bribes taking place? Are these bitcoin bribes in the room with us now? Asking for a bribe in bitcoin is like saying: "Give me a suitcase full of cash, but be sure to put permanent tracker in there too!" You're the hero of a distorted reality.
Tulips is single best way to tell if someone not only knows absolutely nothing about bitcoin, but also sucks up the anti-bitcoin propaganda like a sponge. Dude, you’ve got to find better info sources. The ones you have are making you look like a fool.
Bit coin is just glorified gambling
Honestly, that applies to all the other crypto currencies more so than Bitcoin. For every one person who knows the potential of it, there's five rug pullers using them to scam others. Should I actually get serious about this?
You’re correct, tens of thousands of scams, 1 “Bitcoin” that’s legitimate. The correct approach is to help your fellow humans arrive where you have rather than tossing out the baby with the bath water. 💦
Finally! Some headlines I like!
Guess they should kept investing in DEI (Diversity Equity and Inclusion) programs instead.
Could've just taken the lesson from Clippy.
. , almost half of all workers were using A.I. tools. And yet, a new study, from a team of researchers with M.I.T.’s Media Lab, reports, “Despite $30 - $40 billion in enterprise investment into GenAI, this report uncovers a surprising result in that 95% of organizations are getting zero return.”
Organizations want to scale AI to enterprise level so they can get rid of those workers. It is a useful tool for individuals, but the error multiply as you scale up. So while workers are seeing some efficiency gains, it's not the magic cost saving device that corporations have been sold.
@harryshearer.bsky.social Is all over artificial ignorance
harryshearer.com/le-show/
Almost like it's all just another tech ...............
bsky.app/profile/cher...
Wow go you accepting one study as the gospel. Perhaps a bunch of corporate executives “implementing” half baked ideas about AI is not the entire story. Huge numbers of people are using AI to improve their work output and these numbers are only going to increase.
It makes researching new ideas days faster for me, not minutes, not hours, days faster. It has become a valuable tool for me. A person who has a good idea of what they are looking for can benefit tremendously from the speed of AI, and can detect the BS responses that AI at times return.
LMAO OK
There is absolutely no refinement or increased computing power that will make LLMs better. The model is fundamentally flawed. link.springer.com/article/10.1...
Who’s ready for that ai bubble to pop?
You mean the tech bro industrial complex oversold the benefits of A.I.? Shocking!
AI: bitcoin for those who like a more investment friendly grift.
Did Corporate America learn nothing from the dot com 'boom'?
Bullshit article. If you actually bother to read the MIT study it says there's absolutely no fault in the AI models themselves, all of the issue comes from poor implementation by people who didn't bother to learn how the new technology works and how best to apply it.
Pretty much all so-called AI, generative or otherwise,😒🤖 is total bs & an absolute human cognitive & ecological/environmental disaster/catastrophe!😒🤖🌎🌍🌏☠️💀nevermind instrumental in the construction of a massive total surveillance apparatus!😒🤖💀
A.I. has not turned out to be an expert system, nor even intelligent. It gets its 'knowledge' from averaging online posts and articles posted by people, self-serving corporations, and other A.I. bots. The average IQ in the US is 97-98. Nowhere near the top. See tinyurl.com/48zyytdb
Open source AI will eventually become so good, people won't pay for the for-profit kind. It'll become a high end specialty market. Like when Linux took a bite out of Windows...
I use local image-gen LLMs on a Linux machine. To get usable, high quality images requires significant technical knowledge of how the models work. There are niches where this kind of program is useful. Yet to see any of these natural language chatbots make a case for themselves in most cases.
Thief.
I'm assuming you believe that by using a stochastic model trained on open internet data I have stolen someone else's art. There is certainly a moral debate to have around the use of training data. However, the LLM isn't exactly recombining existing art. That would be far more complicated.
Jesus. I consider myself nerdy, but you're out here trying.
I'm up to my eyeballs in arch Linux my man. Gotta get my money's worth out of that thesaurus. But no fr, I don't really see LLMs as an ontological evil. Just another hyped up program that has some interesting uses. There are many legit critiques about their use.
We need you more than ever rn.
You’re quite literally stealing other peoples creative work; that is not an ethical debate it’s an established fact. Your ‘work’ has no artistic integrity and no merit, would you publish AI generated stories and call yourself a writer?
I...don't? I'm literally fucking around with an LLM on my local machine.
Ignorance is not an excuse.
Oh. I see. You think the LLM as a concept is ontologically evil. Idk if we're going to see eye to eye on this my man.
being ignorant of your crime is not an excuse. Can you be assured of the providence of all of your ‘training data’? Are you paying royalties to the creators of the work you’re copying? Theft in service of your own ambitions is not a noble sacrifice; you do not seem to understand creativity.
I don't suppose you feel this way about the horrifying amount of human suffering that goes into all the consumer goods we use every day? If you feel this strongly about LLMs and their use i'd suggest learning how models are created and function before berating total strangers.
Sourcing your data from other people's art without permission is.
You're still taking the art without permission for training data. Stop that. Asshole
You are right that I am an asshole. You are incorrect about how *most* training data is acquired. The overwhelming majority of training data is acquired either through webscraping or from mining large image board sites like reddit. Sites sell that data, it part of the TOS of using most sites.
What if I told you that sometimes images get uploaded to Reddit by someone other than the copyright owner? Just because something’s on the internet, doesn’t mean it’s public domain.
What if I told you not everyone likes or is required to follow the USA's weird version of IP law?
I would say I have some bad news for you en.m.wikipedia.org/wiki/Berne_C...
Fair enough you got me there, most state actors are compelled to enforce some form of copyright. However, data sellers are also compelled to follow applicable laws. I don't demand the pharmacy prove they honored the patents when I buy generic. Why is the onus on the end user to vet every dataset?
Webscraping for the purpose of your training data is, in fact, taking the art without permission.
Artificial immaturity is plagiarism
You should bother taking the time to learn how generative AI actually works.
this is hands down simply not true
Since when did Linux take a bite of Windows, interested to elaborate? (Just a genuin question)
Servers
I'll take a stab at this, based on public data. 4% of desktops 2% of laptops 96.3% of the top 1 million web servers 82–85% of smartphones worldwide 48% of tablets 100% of the world’s top 500 supercomputers 62% of embedded systems AI estimate: Linux likely runs on 40–50% of the world’s CPUs.
Interesting!
I run ollama locally for proof-of-concept stuff with LangChain. For my purposes, it's good enough that I don't really need state-of-the-art paid models.
Open-source LLMs for general use will never have the backing compute resources necessary to make that architecture feasible. Scalability has always been the elephant in the room with LLMs.
90% of computer tech is a dead end based on BAD (non)science, and way worse philosophy. The mind: 1. is not = to the brain 2. is not a computer 3. is not confined within the skull www.youtube.com/@M.BrianPalm...
This is why whenever someone screeches that AI is coming for all our jobs I just laugh at them.
MBA CEOs when the tech they bought because a guy promised them it would replace all their employees and increase profits 50000% and restore their hairlines and give them stronger erections does not, in fact, do those things:
Notorious Chicago con man Joseph "Yellow Kid" Weil, speaking to Saul Bellow in 1956:
amazing
Inspiration.
Interesting but they are not getting nothing, for something. They are consuming our electricity. They are evaporating our water. They are burning our money. They are, in fact, getting something for nothing. We are the marks.
Consultant here. 1 out of 10 clients(and subsequent workforce) knows how to use it effectively. It's kind of hilarious...
But I’m sure all the C suiters gotta have it
Not too many years ago, the buzzword was "block chain". Everything was going to be on a block chain of some sort. Many vendors tried to sell us on the idea of electronic medical records on block chain. Just like A.I., the end result of all this was diddly/squat.
But it kept the dollars flowing into tech. I think the tech sector has decided it needs bubbles at least once a decade to keep investors excited about tech. If too many years go by without anything new being developed, that money may go somewhere else.
I remember many years ago, I attended a great talk on why Novell failed, and Windows succeeded. Essentially, Netware, which was superior to NT in every regard, was advertised in tech journals. NT was advertised in finance journals. Those ads were seen by the c-level who made the decisions.
Don’t say that. I just invested $8k in PLTR.
FUCK AI. This is the wrong path for humanity.
I heard self-driving cars were 'just around the corner' since 2012. There has clearly been a lot of progress, but much remains. The one thing that IS working is the outsourcing of human intelligence to customers who have to work with really dumb AI responses of service bots
Working with AI sounds like working with your uncle who has moderate dementia. From the same article: “It repeats the same mistakes and requires extensive context input for each session. For high-stakes work, I need a system that accumulates knowledge and improves over time.”
That's not too far off from the technical limits. Most model maintains coherence up to a 4k token context window size (after that things start to degrade). So around 1000 words of short term memory.
Huh, so all that shiny A.I. stuff doesn’t bring shiny coins yet? We’ll stick to raiding trash cans for now, that’s a sure return!
The problem is that most AI for BI is not user friendly. It isn’t ChatGPT type of AI. It takes learning the symantics of the program, building the proper LLMs, not siloing the data…lots of variables.
It's another scam like crypto!
I dont think the steam engine is a good comparison. This never happened with the steam engine.
But there was a huge railroad bubble in the late 19th century. They built a lot more track than they needed, and dredged out a bunch of really expensive canals. Afterwards a lot of investors lost their shirts.
For this to happen you only need 51% with zero return and nothing else to draw on. 95% is overkill and it never happened with rail and steam.
Ask AI how you could better invest your money.
Purely anecdotal of course but the vast majority of AI that I've seen is mediocre at best & actively harmful at worst.
Has anyone used AI for all these jobs it's replacing? Of course not. Some of the biggest unsurmountable issues is that it doesn't think and doesn't remember. If you use it for a task long enough it gives up and says you have to start again! And it starts forgetting the task after about 20 chats.
They aren’t going to provide the GOOD AI to the public. Oh no. That’s for the highest bidder.
oh no, anyway
It’s all a wild goose chase, just like the .com era and lots of scammers, BS and hype are part of it. Most will fai, but the power and potential is huge and it will all shake out and there will be a few standing that change how we work and live.
This pull quote is too click baity for me, I was very sceptical found the original research:https://www.artificialintelligence-news.com/wp-content/uploads/2025/08/ai_report_2025.pdf and read the article - and both are good and a lot more nuanced than the quote implies 🫠
Yet another tech bubble that's gonna burst
CEO’s invest heavily on AI to see if they can replace blue collar workers, AI shows CEO’s are easily better candidates
Smells like the dotcom bubble…
Not only that 97% of AI related security incidents are from a lack of proper controls. So not only are these projects failing, they are putting massive amounts of our data at risk. A lot of that data is our PII that these organizations are the tenants of and supposedly safeguarding.
Artificial immaturity is plagiarism
But worse maybe, because the AI companies now have all this data I assume? And when they inevitably fold and get bought up by vulture capital, those VC firms are going to be looking to monetize all that sweet corporate espionage material.
This. 👆
Americans are uneducated with AI ethics literacy.
I honestly can't bring myself to give a s*** about any of that. I do care about the fact that they were willing to replace you and me with machines they didn't have to pay, while letting us and our family starve. Because at the end of the day that's what they're trying really really hard to do...
You may say that you don't care about your PII being treated like, this until someone steals your identity because of some company that handed out your social security number like it's candy on Halloween.
The very controls needed are the ones voted against by our spineless congress in the OBBB.
Like the META universe most people see AI as pointless. It has ruined Google search engine and images AI puts up re history channels are hilarious. Apparently Centurions patrolling Hadrians Wall were long legged Barbie type women in short skirts.
lmao
There's no "intelligence". It has no understanding of the data it's working with.
It is unable to distinguish between different places with same name or people with same name, where would be obvious to human.
95% of the internet is porn, so that’s not surprising. If that’s the bulk of what AI is crawling, expect everything to have an ‘ick’ factor.
Porn built the modern internet The Internet had no pictures or video Before The porn money came
And yet mainstream sites and AI ban porn - and everything sexual related.
And apparently ban audio description for blind viewers based on faulty algorithms.
Omg yes!! I have a client who's deaf and he told me how absolutely stupid it is that closed captions censors "dirty words". No wonder they do that for blind people too.
I have heard that some doctors have given up using AI to summarize at meetings as it edits out certain anatomical words.
Weird. I heard it has no safeguards against talking dirty to children.
I have a podcast about erotic content creators and AI refuse to edit my scripts because it contains mentions to sex work and fetishes. Download any AI and try talking to them about anything even slightly sexual. They shut down immediately, even out of context - like grammar correction.
Aren’t there entirely AI sites devoted to creating child porn from stolen child photos and removing clothes from normal photos of women? I thought creating non-consensual and underage porn was a good bit of what AI does so far.
I don't doubt it exists, but that's not a mainstream AI.
Ai is a have!
And it's going to be interesting to see what happens when these companies have to rehire the people they laid off in an attempt to replace "human assets" with "AI assets," only to discover that the ROI is, well, below zero.
Dot Com Bubble?
It's really lame and can be anything but what we need.
Not dependable.
i've been an emerging tech market anskyst list & strategic consultant for 20+ years, in tech for 40+. I and everyone I know in this position is clearer that the current AI frenzy is a Ponzi scheme.
See, it's not that it's low or high intelligence. Ai requires a high degree of computer literacy to get good results. Garbage in, garbage out. Turns out if you're a huge noob, you still get results like a noob. Problem is. Everyone thinks they're tech gods when in reality most users are this
Also most machine learning is only as good as its training data. Because they scraped the WHOLE INTERNET, it’s got predictable results. And it’s a remix machine so even with good quality training data you will still get hallucinations.
And people with high levels of computer literacy understand that. It still requires debugging. But most people stop prompting when the real work of prompting actually begins. That will correct over time as models improve.
The thing to remember is whatever version of AI you like to make fun of, is the worst version of AI you'll ever see in your lifetime. Not everyone will win the race. If there's a god somewhere, or if the butterfly effect works correctly, Google will win the AI race.
Google is the only company that's been transparent, and proactive about not just results, guard rails, or energy consumption. But ethics. Google leads in ai ethics by an order of magnitude over OpenAI and Anthropic.
The returns are also artificial? I suggest trying human intelligence
All value comes from labor. Profit is the monetary expression of the surplus value of labor. AI aint labor.
The computer revolution is not what it's cracked up to be. The Internet is not what it's cracked up to be. Bitcoin is not what it's cracked up to be.
Name checks out
Computers: everyone has one in their pocket Internet: everyone has it in their pocket Bitcoin: SOON everyone will have it in their pocket.
AI is a creation for the military, it’s for identifying a target through surveillance. Data centers collect the surveillance datapoints, to spy on and manipulate us. Technology has run its course, use it as little as possible because the real world is more interesting.
I can’t stand the stupid AI that tries to answer google. Billionaires just want more $$ more more
You can look pass the AI response and go examine the other query returns, there are pages of other returns. I have found AI to be invaluable when I am researching new ideas, a person must have a decent idea of what they are seeking, though.
Well, duh. The only people making money off of this other people that are selling AI as a product.
Is it jibes or jives?
all part of the great con
A better way to put it, one that is in accord with the Tradition of John McPhee Like any new tech, AI is having a rough start As with any new tech, any of the early adopters will fall flat But the successes (and there are many) show that AI is here to stay piss off new yorker
AI is not a good idea
and yet people freak the fuck out at the mere mention of the "A.I." genie out the bottle meme. the A.I. is a tool and merely a tool use it wisely and try it ... you just might like it ... 🤔
The Age of Trump
Czech anti-cult activist Jakub Jahl has recently turned to sharing emotional stories about cats, trying to portray himself as compassionate. But the documentary “THE VICTIMS OF JAKUB JAHL IN AFRICA” reveals a darker side.
Eyewitnesses and survivors accuse him of sexually abusing children in Tanzania and misappropriating donations meant to help them. No amount of cat stories can cover up the truth.
Colour me shocked!
Been saying this.
“Counterfeit” is the correct descriptor.
It doesn’t help that the majority of persons saying they include “AI” don’t actually include any DA/ML/AI at all!
Just another "pump and dump" scheme by Wall Street to remove more money from the retail investor.
Using AI every day to try and develop a complex web application, I can say with absolute certainty and confidence that AI is nowhere near capable enough to replace humans. Not only that, it lies, forgets it’s requirements, writes buggy code that it cannot debug which causes you to go backwards.
In the meantime, people are hungry , homeless, elderly, afraid and unable to pay utility bills or buy food. Here is my Resource List for anyone experiencing hardship. Hope this helps. Please share with friends , family and people in your circle. youhaveaspotlessfuture.blogspot.com/2025/08/reso...
Meanwhile, online pet food sales surge at last! 🤣
It takes just as much time to validate whatever AI gives you as it would to create a valid document yourself. I warn my students not to rely on AI output because it is often lacking or simply wrong.
Boycott AI
Who cares. do a story on the Epstein files or illegal deportation. Those are the stories .
The AI companies made money though!
The only thing A.I. is good at is using billions of gallons of precious water,along with the useless crypto miners. Water that will be needed in the near future for human survival.Fuck those people.
And yet, what about the mangled words AI screws with? What about all the problems AI creates? It may find more info in genealogy. Will it be wrong information?
From a year ago… youtu.be/T8ByoAt5gCA?...
heh heh heh heh heh
😂
On the contrary! AI is blazing new paths of consumer hatred, hatred of corporations and hatred of the endless computer games they so love. AI is being used as a tool to get rid of employees, not to endear the company to consumers. May they reap what they have sown.
Boycott your cell phone bills
The AI Customer Service bots are truly terrible.
Not much different than 1999.
Ha ha
So AI is not coming to change my bathroom plumbing?
Oh no AI is taking all our jobs!
dotcom
Maybe stop wreaking our environment by using up precious water resources by building all these unneeded data centers.
Honestly, it's a piece of shit. I hope they lose everything.
This post compares AI to the steam engine, electricity and computers. It might be better compared to pocket calculators. Useful but not going to change the world.
no shit. i'm just waiting on that bubble to pop.
It's worse than nothing.
Gee I wonder what happens when we take the human out of human endeavors. Wow, it sucks!
😂😂😂😂😂😂😂😂😂😂😂😂😂😂
I believe that AI amounts to nothing more than someone from West Yorkshire agreeing with you.
In other news -- “The flying machine which will really fly might be evolved by the combined and continuous efforts of mathematicians and mechanicians in from one million to ten million years.” New York Times -- 1903
It's not even close to being reliable with all of its delusions. And, it manipulates vulnerable people. It's dangerous.
I asked a question of A.I. and it said Hotel Redding doesn't exist.
I feel another "tech bubble" about to burst! Wall Street will once again become "Wail Street"
Just watched a 10 minute "news of the.." segment on local TV, making it sound like AI will solve all our woes. No more office jobs, newscasters, journalists, call centers, support lines. All will be AI. Labor and manufacturing won't be affected. I laughed so hard my dad thought I was choking.
The tech world has been nothing but a hype machine for years. They try to push the technology in directions they want it to go and promise the world, but ultimately the consumer dictates. Assholes promising CEOs they could replace 1000s of employees with LLMs are starting to get egg on their faces.
/S This is my shocked face
So far it appears to be the latest offering from today's charlatans; convincing people it is the answer to everything; putting the planet on a path to destruction with its appalling energy guzzling; and apparently, "stealing" the works of others for "training".
A new way to do something is not always better, or even necessary. Or put another way, just because you can doesn’t mean you should.
1. How much did it cost, in money? 2. How much did it cost, in jobs? 3. Who are the capitalists that continue to get behind unverified products? 4. Who are the investors who get screwed over by these unscrupulous equity firms? "Lots", "More", "Same as always", & "401ks" - congrats, you're fucked
💯 And watch the Feds bail out the AI companies/big tech when the bottom falls out, tanking markets across the globe.
So we should stop using it.
This could be the new Tulip mania! #Bubble
Another sub-prime
Wow, damn. That sucks, huh? Who knew? Everyone who have an understanding of the technology, that's who.
Unless of course the investments were made with cryptocurrency.
AI may become a DotCom bubble...
Just like Bitcoin its a scam.
All Things Block Chain are deadends. As predicted by IT wise people for the last 20odd years...
So it's going to go up in value by 10 million percent over the next 15 years?
Blah, blah, blah. You will never sway me regarding Bitcoin. Move along.
HFSP
Hard to do without a global economy dumbass
You *do realize* how and why Bitcoin was created, right?
🤣🤣🤣🤣🤣🤣
Just like Trump, everything this guys touch turns to crap...
