But surely this amazingly useful product that definitely 100% works must be making so much money on its own. Right?
I wonder how long google would last if everyone in the world would just not be online? turn off your phone. don’t use your PC.
what would it take to starve google of the data it’s dependant on? a week? two weeks? a month?
how much would it take to irrecoverably damage their business?
I wonder how many other businesses would be damaged at the same time?
I had an uncle who had his life quite on track. Good job, nice house, beautiful wife. Than he hit midlife crisis and learned about online gambling. He eventually lost everything, but it was a slow, painful, and much too public process. He died not much later. Completely alone and in the gutter.
This comment ^ is about the “AI” bubble
nah don’t worry. they will pay them out from our pensions.
We can only hope that’s what happens here.
I hope all 80bn of it become obsolete and won’t be useful within 1 month.
Surely there has to be a level of expenditure where they just can’t reasonably make this back, right?
Like the debt these hyperscalers are in to their investors at this moment surpasses the debt of entire countries, yet they somehow expect to make this all back and then some??
I think the major investors and CEOs all privately know this is insanity but are trying to capitalize on the wild speculation as much as possible so they are publicly saying otherwise. They probably think that they will have insider information about the bubble bursting and can get out in time, and they’re probably mostly right. It’s the normies with 401ks that will be left holding the bag.
Or, they are hoping that the bubble bursts while the Trump crime family is in charge and that big tech bribes will result in big tech bailouts. And they’re probably right about that too.
So far the money they put in the hole has come back doubled every time. It’s an old scam, sooner or later if just doesn’t come back at all.
I mean this without trying to sound facetious at all, I think a lot of the decision makers in these companies that are pushing so hard because they believe this is their path to godhood/immortality, and no price is too great for that in their eyes
I think that may be true for Peter Thiel or Larry Ellison, but for the rest of them it’s just the same old game: riding the wave of whatever investors want to see. The actual business only matters as far as they can use it to convince investors to keep throwing money at them.
If you gave them a button that would give them a million dollars but a child would die, they’d press it without a thought. They’d press the button until there were no more children on the planet. They’d use the money to convince people to have children just so they could keep pressing the button.
IMO most of these CEOs are so excited about AI because it’s just been the best money printer in a while. Investors are irrationally interested in AI, so these CEOs will continue to find any excuse to say the letters “AI” at least until people stop sending them money for it.
It cost about US$36B to run fibre optic cable to almost every house in Australia.
WTF kind of AI project costs twice that?! They’ll have to claw back around $250 per US citizen to break even.
Have you seen DRAM prices? That’s like 5 sticks.
The datacenters are absolutely paying less than that. If not directly, then through kickbacks from the supplier. The inflated price only applies to us.
The datacenters are absolutely paying less than that. If not directly, then through kickbacks from the supplier. The inflated price only applies to us.
The datacenters are paying less than $16 Billion per stick of DRAM? You sure?
It begins
They’re funding this with shares of the company. Even though this investment will flop, they still have income as the number 1 data broker and ad company on the planet. Their investors will be unhappy to see the number go down, but it’s hard to bankrupt a company as large as Google.
I’m sure Alphabet is up for the challenge.
This can’t crash and burn soon enough. Might be time to set fire to some datacentres
This doesn’t seem like a good idea.
This is a fucking joke.
There is so much over investment, they could just wait a few months and buy all this shit for so much once openai craters.
I know what I’m not buying.
I hope that Micron, SK Hynix, and Samsung don’t hit any hiccups with their memory chip production buildouts.
I thought they weren’t building it out because it’s more profitable for them when there’s a shortage?
They’re building out. The first ones are going to be mid-late 2027, but those aren’t expected to be at full production until about 2028.
searches
Hmm. Micron says that they’re aiming to move up their first fab becoming active to the first half of 2027, now:
https://www.idahostatesman.com/news/business/article314253330.html
Micron CEO Sanjay Mehrotra says the company is pushing up its timelines for opening its first semiconductor fabrication plant, or fab, into the first half of 2027, instead of the second half. His announcement in a Dec. 17 earnings call came months after the chipmaker unveiled plans to start construction on a second fab in 2026, eyeing a 2028 ribbon-cutting.
They also have a second fab in Boise, Idaho, and some in New York that theylre building.
Both Samsung and SK Hynix have fabs in South Korea that they’re building.
