This sounds like the VC money is running out and the subsidies will soon end.
Don’t you worry, the AI companies already gobbled up the entire Internets worth of data, every conversation they could find on their social media networks to train their machine.
And just like Uber, you’ll now have to pay a premium to use their service when it finally works. Don’t worry they’ll still give you the stupid auto complete chat bot and tell you that’s the real AI, The real AI works for the corporate captured government and is figuring out how to get rid of you. Their goal is not enlightenment of the human race, It’s depopulation and control.
At some point they will have to monetize and that’s going to be a bad day for everyone.
Imagine having halved your workforce, suddenly you have to pay for compute at what it actually costs.
If they don’t find a new thing to sell to their shareholders they will have to monetize.
And good luck rebuilding all the institutional knowledge that you threw away by firing half your staff, while your former employees are getting rounded up into homeless work camps…
Well, the key word is “right now”. AI companies are gambling that their runway is long enough to reach monetization.
Womp womp
In shadow of doubt that many companies have painted themselves into a corner and will find the ai revenue correction painful. That said, I’m surprised to hear its already more expensive than people without reference to maintenance debt costs. Makes me wonder if there isn’t some good old fashioned malicious compliance at work too…
I built my little empire out of some crazy garbage called the blood of the exploited working class.
I’m just going to note how giga-fucked it is that AI is openly being criticized because it hasn’t led to more layoffs.
Just want to point out the “more layoffs”. We are still in the Uber-subsidized part of the relationship.
This is the honeymoon and it ain’t off to a great start
We are still in the Uber-subsidized part of the relationship.
It’s not comparable, the costs for Uber are fixed development costs which decrease per customer with more customers. AI is strictly more expensive the more users there are since the cost is per use. The financial outlook for the industry is very bad, it will probably never be profitable except maybe in some extremely niche situations.
AI is strictly more expensive the more users there are since the cost is per use.
Isn’t most of the cost in the model training?
The cost of inference has passed the cost of training quite a while ago and it’s by far the majority of expenses
I don’t think they mean uber the company, I think they meant uber as in over or very.
I think they mean the company since they’re the most famous example of subsidising the product to gain market share and their name is invoked all the time in discussions of AI economics
The companies need to at least seem to grow forever. See Facebook to meta VR stuff to AI. The business models could not grow forever that’s the reason for this constant reinvention of themselves.
At some point AI and building these data centers will have to lead to quantum break , wow we have AGI or to monetization.
Who’s going to pay the actual energy costs+ margin for all their tokens? Not me
It also can’t do any job I know of. Weird framing. Though, I guess that’s the snake oil they’re trying to sell.
If you have someone good enough shepherding it, you can get decent results quickly.
It doesn’t turn less money into more output, it turns more money into less time. And, you know, steals from everyone and eats our resources. omnomnomnom
It can’t fully displace jobs because it’s super unreliable and makes tons of errors all the time. But it can do some tasks well if you know how to handle it.
Still not worth anywhere near what it costs, either to the user, the CAPEX investment, society or the planet.
Not to mention the risk we take by running them but being unable to control the things. Because, you know, we don’t actually know how they work, since we didn’t program them, to begin with.
Just imagine where we would be 5-10-15 years from now, if all that AI money would have went into social and environmental projects.
That describes my entire life since adolescence… wondering what kind of society we could have had if we werent governed by sociopaths
If all the resources we spent blowing shit up and killing innocent people had been put to good use instead, we would be in a Star Trek-esque post-scarcity society by now. Possibly on our way to being Kardashev type 1 society.
Slop software houses are gonna PAY BIG for losing human talent.
When that happens and it all crashes down, there will be plenty of experienced devs and sysadmins to fill the market gaps. Maybe as independent contractors/freelancers, but ideally as tech co-ops.
I’m hoping this insane future booked shortage causes the consumer electronics industry to crash due to lack of parts which in turn should cause the AI industry to crash when no one is buying new tech nor fat AI subscriptions.
It already has to be affecting small to medium businesses significantly when even laptop procurement has tripled in price and you’re spending a ton of money for enterprise AI access.
These companies want to own all the hardware so you are forced to rent a computer. To buy a PC that does nothing by itself and requires some remote hardware to function. Amazon is already preparing a “game streaming” service, for example.
Good luck with that to them. Internet reliability is extremely crappy in the US, even on supposedly “tech hubs” in California the best you can do is unreliable Comcast or super slow DSL from AT&T.
I don’t think any of them are terribly concerned with the quality of the service.
Ahh, late to market in a flooded field with at least one law of physics preventing it from ever working acceptably. Good job, Jeff!
I think they’re hoping the result is that we can only afford cheap tablet devices that act as dumb terminals for their cloud services which we have to rent forever due to holding our data hostage so they can manufacture consent and prevent us from developing open solutions to their walled garden proprietary products.
A precarious moment at the edge of a cliff.
It already has to be affecting small to medium businesses significantly…
I’d imagine that is a major factor in why it’s so accepted right now. Gotta consolidate the market!
So… why the fuck are you going all-in on AI at the expense of literally everyone and everything else!?
Control, propaganda, surveillance, tools of war, etc and so forth.
the investors, anyway
They’re nvidia, they don’t care if the shovels they’re selling won’t dig any gold
Short term? Drive wages down with the threat of firing people and replacing them with AI. Long term? They’re either delusional enough to believe AI will improve to the point where it actually is cheaper, or else they’re willing to pay more for a workforce that can’t organize and protest and that they don’t have to worry about doing things like being a whistleblower for their latest amoral plan.
Some CEO’s thought they could save a bundle and were eager to replace workers. Other CEO’s saw that, had to do a bunch of layoffs and knew saying “replaced with AI” instead of layoffs keeps stock prices high. Then you have a whole bunch of CEO’s that thought the other guys must be on to something, think they’re missing out and jump into AI both feet first. The laggards are bombarded with news and propaganda by the big players to get them pulled in too. So by now you get everybody messing around with AI in some fashion. That’s the moment the tech giants put the squeeze on everybody.
They are clearly anti-human. The question is what are they really trying to achieve? I hope it is not just Bioshock.
My theory is that they think that, with world-scale financing, testing, and iterating, they can get this thing to do a lot of work that is currently exclusively in the purview of humans today. I believe that some of the wealthiest among us tolerate the rest of us sharing their air because, for now, they must, and the second they can replace us with obedient machines, they will.
If I am right about that, then this entire hype cycle and seemingly endless funding rounds and unabated lending Ouroboros Ponzi schemes makes more sense. They’re not delusional that the product works better than it does, they are desperate for the holy grail of greed: not having to ever share anything with anyone ever again.
profit. the profit must flow.
…but there is no profit…eg testla
Just wait till executives realize they can’t use AI as a scapegoat when they force a bad decision through the company. They’ll have no one to take the fall.
It’s like that twilight zone episode. They fire everyone then no one is left to help do anything
AI businesses are already getting away with their software killing innocent people. I think everyone will work out for the people at the top.
Yup. Capitalism only works in a finite world if there’s a catastrophic economic hardship for 99% of the population every 5 to 10 years just to ensure the working class doesn’t achieve too many gains. Ai causing the economy to crash is bonus feature, not a bug.
It’s why there’s often a human in the loop who has to check the result of the AI and can be blamed if things go wrong. The poor sap can’t possibly check and correct the volumes of slop produced so they’re just there to work as a scapegoat.
This reminds me of Barney Stinson
Except he was paid well
I’m pretty certain that companies and people have been using the old “computer error” excuse since at least the 1980s.
Are you sure? So far it seems like they’ve been able to do just that.
Gruen tonight was talking about ‘Single Person Billion Dollar Brands’, it was sickening.
Wow!
Really?!
No way!!!
Its almost like anyone who could do grade school math could have pointed that out 2, 3, 4 years ago!
I wonder if anyone did…
Anyway, uh, chain all these fucking morons into each other, SAW style, tell em they all have to work together and mutually suffer to escape the trap.
That’s not murder, it’s comedy.
Not too mention Über and other new services clearly worked on getting people hooked, then jacking up the price.
SAW style
So, their arms are stuck in a device that allows them to reach a keyboard. If they can get the AI to do a task correctly they’re free, if not the device cuts of the arms.
So, their arms are stuck in a device that allows them to reach a keyboard. If they can get the AI to do a task correctly they’re free, if not the device cuts of the arms.
Hah! That’s very good as well, I was thinking something like… they all have a chain going literally through their shoulders, around/between bones, with lock mechanism inside their bodies.
They each have a key, but the way the… dungeon/trap is set up, they can only unlock the chain if they all simultaneously turn the right key, in the right lock, for the right person, all at the same time.
If any one of them dies of blood loss, well, now they’ll all die.
If they break their collarbones, to pull the chain/lock out… I mean good luck trying, without severing your suprascapular artery.
At its current pace, AI expenditures may reach $5.2 trillion by 2030
I can’t wait.
They think it is a lottery and sinking more money into it will increase the odds of “winning at everything”.
I love that their moat has been bridged by open models as soon as it was dug. At the current rate, you just have to wait six or nine months and open models will be at opus 4.6 level which is really all you need for most applications. After this I don’t see how the big labs could ever recoup their losses.
Obviously, but think of all the middle management, payroll, HR, etc you can fire/not hire!
I’m not sure who to root for here. Middle Managers can suck eggs.
Their not firing middle managers. Their firing front line coders and lower desk employees and keeping the middle managers while using AI.
They’re *
Theiyr*re
That’s not really what’s happening they’ve been trimming a lot of middle manager fat too
