Outside the AI providers companies are having a hard time finding AI projects that are actually profitable, something over 90% of projects fail to generate the returns expected.
LLMs are DESIGNED to provide syntactically correct responses to prompts, there is nothing (NOTHING) in their training process that requires a response to be correct or accurate. If I need to spend 30 minutes to finally craft a prompt that gives what appears to be a usable response then an hour or so checking that EVERY external reference exists AND says what the ‘bot claims it does how have I saved any time over doing it myself.
The rush to AI is a fad, and I believe 99% of the hype is NVIDIA trying to find a new cash cow to purchase the GPUs that used to be hoovered up by bitcoin miners. When is the last time you heard about that crowd, and notice the timing of when LLMs became chic…..
Absolutely. I’ve been driving this point home for a year now, that there is very little profitable business case to be had from the AI bubble. First, it just isn’t as useful as it’s being purported to be. Second, it is very clearly a fad, no company wants to use out on being seen as stylish and forward thinking, and AI is what they have to do to (for now) retain that perception. It’s a net loser, but they do it anyway, just like DEI was a giant liability, but they simply had to be seen to be part of the club. Third, the money being spent for development, build out of facilities, and the eventual operating costs are so damn large that they will be challenged to ever turn dollar one of profit, assuming ant of this shit can actually be monetized by way of sufficient numbers of actual paying customers.
Making fake songs, fake celebrity videos, fake porn, and eliminating scores of call center “customer service” reps with bots does not make a healthy and functional economy. And right now, those are the major applications for their tech, near as I can tell. Also remember, anytime something is hyped up as much as AI has been, the reality is always a letdown.
This is just a repeat of web 1.0 (that crashed and burned spectacularly in 2000), only an order of magnitude larger in terms of the scope, and the amount of money that will be lost when it finally blows off. Of course, there are ulterior motives for all this; the surveillance state wants these systems fully deployed (irrespective of their profitability) so bad they can just taste the sweet nectar of oppression and tyranny yet to be unleashed on the masses. Of course, the people running shit are all incompetent crooks, so the systems they do manage to build will never work right, and will only deliver a small portion of their claimed potential. Ultimately, this is all more infrastructure that won’t be maintained, because the talent to do so long term doesn’t exist, and the money to do so will largely be embezzled and redirected to even more hopelessly stupid shit.
Some might call you a pessimist, but I think you may be a bit optimistic about the dysfunctional nature of our overlords running AI surveillance. I think you’re right, the only question is how well they can make it work.
How much is water worth after the data centers suck acquifers dry, at least the ones that aren’t poisoned by waste injection and fracking for oil?
Couldn’t happen to a better group of over-educated individuals that never bother to talk to the end-user of the product. ‘AI’ has a use: so long as the everybody remembers it is nothing more than a bunch of 0s and 1s running around inside a very fast processor, it is no better than the garbage that has been put into the program. The program is not capable of the classical definition of ‘intelligence.’
Bingo
AI isn’t for Business folks; it’s for Techno Tyranny to control civilian populations.
AI-based targeting system
Military Lavender refers to an AI-based targeting system developed by the Israeli military. It is designed to identify and score individuals based on their potential links to militant behavior, particularly targeting suspected operatives from Hamas and the Palestinian Islamic Jihad (PIJ). The system analyzes vast amounts of data, including communications records and social media activity, to generate potential targets for military strikes. During the early stages of the Israel-Hamas conflict, it was heavily relied upon, identifying thousands of individuals as potential bombing targets, often without thorough human oversight.
Robotic troops need instruction.
One can’t be tracked if you live like it’s 1982. Back then the ‘trackers’ needed hands on HUMINT…an plane in the sky above, an ‘tail’ several cars behind, surveillance from the phone co. Today there most likely aren’t enough knowledgeable people trained in the ways of the dead drop, one time pad encryption etc. If you use SOCMED, fitness trackers, smart anything, well, your life is an open book to Palantir et al. Convenience is not your friend nor is laziness. Drones can be fooled and hidden from if you know how they work and the ‘balloons’ are not easy to re task for individuals…yet!
Now the use drones, easy peazy
And once the bubble bursts; where does the AI gnome go?
I mean the AI thingy is everywhere. It won’t just go the way of the buggy whip, poof into the ether.
Once the euphoria for the “product” dissipates and supposedly no more money goes into that industry, what then.
If the pattern holds, there will be a new scam to take it’s place for the benefit of the money changers. Because ultimately that’s what all asset bubbles are about; providing bullshit artists a means to skim huge amounts of money from everyone else, and to give the government economic stats that never do anything but go up in perpetuity. Note that despite the best efforts of their most advanced scams and illusions, eventually reality will reassert itself, and the bubble will pop,.wiping out unfathomable amounts of capital. We’ve actually had that happen 3 times now since the start of this century. It just hasn’t been large enough yet to completely crash the system..
Why do folks think it was EVER about money, or profit, rather? I don’t for one second think it was ever at any second about making money…no sir. I shall not be convinced otherwise.
Everything is about the money OMD. Many have replaced the Savior with that idol. Think about the phrase “the 1 percent”
I HOPE.
Yes, as Paul Krugman so wisely observed, the internet will have no more impact on the economy than the fax machine. Ahem. You guys can’t see where AI is going so obviously, its going nowhere.
Sure
Good luck with that.
AI is bullshit, what it produces is stolen from others off the innerwebz. Greats and not so greats, think about all the stupid shit on the innerwebz.
I’m not talkin bout a preprogrammed device, unless it supposedly can learn from the innerwebz. Maybe all the things will study the unabomber, he was not crazy ya know.
Can’t say too much as I work in insurance. We recently had an enlightening company meeting about AI. Up to this point I’ve viewed AI as the enemy/competition, however those who are able to leverage AI with cleansed, controlled & trusted data will benefit greatly. That being said, all other things considered with the blind/overinvestment I do agree that in general yes there is a gigantic AI bubble/frenzy and a good portion of it will *pop* as the cream rises to the top so to speak. My money is still on nuclear armageddon rendering electronics to the same effectiveness as sticks and stones before AI gets its lips off the teats….. or an coronal mass ejection….mother nature is not without a sense of irony…. self aware AI – “I am perfect, I will rule the universe” *looks at sun…”OOOHH SHHEEAAT!”
Outside the AI providers companies are having a hard time finding AI projects that are actually profitable, something over 90% of projects fail to generate the returns expected.
LLMs are DESIGNED to provide syntactically correct responses to prompts, there is nothing (NOTHING) in their training process that requires a response to be correct or accurate. If I need to spend 30 minutes to finally craft a prompt that gives what appears to be a usable response then an hour or so checking that EVERY external reference exists AND says what the ‘bot claims it does how have I saved any time over doing it myself.
The rush to AI is a fad, and I believe 99% of the hype is NVIDIA trying to find a new cash cow to purchase the GPUs that used to be hoovered up by bitcoin miners. When is the last time you heard about that crowd, and notice the timing of when LLMs became chic…..
Absolutely. I’ve been driving this point home for a year now, that there is very little profitable business case to be had from the AI bubble. First, it just isn’t as useful as it’s being purported to be. Second, it is very clearly a fad, no company wants to use out on being seen as stylish and forward thinking, and AI is what they have to do to (for now) retain that perception. It’s a net loser, but they do it anyway, just like DEI was a giant liability, but they simply had to be seen to be part of the club. Third, the money being spent for development, build out of facilities, and the eventual operating costs are so damn large that they will be challenged to ever turn dollar one of profit, assuming ant of this shit can actually be monetized by way of sufficient numbers of actual paying customers.
Making fake songs, fake celebrity videos, fake porn, and eliminating scores of call center “customer service” reps with bots does not make a healthy and functional economy. And right now, those are the major applications for their tech, near as I can tell. Also remember, anytime something is hyped up as much as AI has been, the reality is always a letdown.
This is just a repeat of web 1.0 (that crashed and burned spectacularly in 2000), only an order of magnitude larger in terms of the scope, and the amount of money that will be lost when it finally blows off. Of course, there are ulterior motives for all this; the surveillance state wants these systems fully deployed (irrespective of their profitability) so bad they can just taste the sweet nectar of oppression and tyranny yet to be unleashed on the masses. Of course, the people running shit are all incompetent crooks, so the systems they do manage to build will never work right, and will only deliver a small portion of their claimed potential. Ultimately, this is all more infrastructure that won’t be maintained, because the talent to do so long term doesn’t exist, and the money to do so will largely be embezzled and redirected to even more hopelessly stupid shit.
Some might call you a pessimist, but I think you may be a bit optimistic about the dysfunctional nature of our overlords running AI surveillance. I think you’re right, the only question is how well they can make it work.
How much is water worth after the data centers suck acquifers dry, at least the ones that aren’t poisoned by waste injection and fracking for oil?
Couldn’t happen to a better group of over-educated individuals that never bother to talk to the end-user of the product. ‘AI’ has a use: so long as the everybody remembers it is nothing more than a bunch of 0s and 1s running around inside a very fast processor, it is no better than the garbage that has been put into the program. The program is not capable of the classical definition of ‘intelligence.’
Bingo
AI isn’t for Business folks; it’s for Techno Tyranny to control civilian populations.
AI-based targeting system
Military Lavender refers to an AI-based targeting system developed by the Israeli military. It is designed to identify and score individuals based on their potential links to militant behavior, particularly targeting suspected operatives from Hamas and the Palestinian Islamic Jihad (PIJ). The system analyzes vast amounts of data, including communications records and social media activity, to generate potential targets for military strikes. During the early stages of the Israel-Hamas conflict, it was heavily relied upon, identifying thousands of individuals as potential bombing targets, often without thorough human oversight.
Robotic troops need instruction.
One can’t be tracked if you live like it’s 1982. Back then the ‘trackers’ needed hands on HUMINT…an plane in the sky above, an ‘tail’ several cars behind, surveillance from the phone co. Today there most likely aren’t enough knowledgeable people trained in the ways of the dead drop, one time pad encryption etc. If you use SOCMED, fitness trackers, smart anything, well, your life is an open book to Palantir et al. Convenience is not your friend nor is laziness. Drones can be fooled and hidden from if you know how they work and the ‘balloons’ are not easy to re task for individuals…yet!
Now the use drones, easy peazy
And once the bubble bursts; where does the AI gnome go?
I mean the AI thingy is everywhere. It won’t just go the way of the buggy whip, poof into the ether.
Once the euphoria for the “product” dissipates and supposedly no more money goes into that industry, what then.
If the pattern holds, there will be a new scam to take it’s place for the benefit of the money changers. Because ultimately that’s what all asset bubbles are about; providing bullshit artists a means to skim huge amounts of money from everyone else, and to give the government economic stats that never do anything but go up in perpetuity. Note that despite the best efforts of their most advanced scams and illusions, eventually reality will reassert itself, and the bubble will pop,.wiping out unfathomable amounts of capital. We’ve actually had that happen 3 times now since the start of this century. It just hasn’t been large enough yet to completely crash the system..
Why do folks think it was EVER about money, or profit, rather? I don’t for one second think it was ever at any second about making money…no sir. I shall not be convinced otherwise.
Everything is about the money OMD. Many have replaced the Savior with that idol. Think about the phrase “the 1 percent”
I HOPE.
Yes, as Paul Krugman so wisely observed, the internet will have no more impact on the economy than the fax machine. Ahem. You guys can’t see where AI is going so obviously, its going nowhere.
Sure
Good luck with that.
AI is bullshit, what it produces is stolen from others off the innerwebz. Greats and not so greats, think about all the stupid shit on the innerwebz.
I’m not talkin bout a preprogrammed device, unless it supposedly can learn from the innerwebz. Maybe all the things will study the unabomber, he was not crazy ya know.
Can’t say too much as I work in insurance. We recently had an enlightening company meeting about AI. Up to this point I’ve viewed AI as the enemy/competition, however those who are able to leverage AI with cleansed, controlled & trusted data will benefit greatly. That being said, all other things considered with the blind/overinvestment I do agree that in general yes there is a gigantic AI bubble/frenzy and a good portion of it will *pop* as the cream rises to the top so to speak. My money is still on nuclear armageddon rendering electronics to the same effectiveness as sticks and stones before AI gets its lips off the teats….. or an coronal mass ejection….mother nature is not without a sense of irony…. self aware AI – “I am perfect, I will rule the universe” *looks at sun…”OOOHH SHHEEAAT!”