OpenAI is expecting to reach $278,000,000,000 in negative cash flow by 2030, putting them over a quarter of a trillion dollars in the red to build AI infrastructure.
They're still trying to figure out how to make AI more profitable.
They're still trying to figure out how to make AI more profitable.
Sep 20, 2026 路 5:00 AM UTC
86
113
1,509
200,275
@pubity Waste of money
Waste of water
Waste of electricity
Waste of resources
All that for negative impacts on both humanity and the environment.
AI is the most expensive mistake ever made, and ever pushed.
Waste of water
Waste of electricity
Waste of resources
All that for negative impacts on both humanity and the environment.
AI is the most expensive mistake ever made, and ever pushed.
4
26
311
4,796
@pubitynomics look even crazier when you consider how much frontier labs are spending on compute while open models keep getting cheaper and are already good enough for a lot of specific use cases
hard to justify that level of spending if the gap people are willing to pay for keeps getting smaller
hard to justify that level of spending if the gap people are willing to pay for keeps getting smaller
0
0
4
712
@pubity Maybe fucking get RID of sam altman?
The board tried to remove him once, because he was a danger, and then he was reinstated and the board dissolved
Hes a loose cannon with no checks in his power in the company.
Hes also a sociopath that thinks AI is more important than us
The board tried to remove him once, because he was a danger, and then he was reinstated and the board dissolved
Hes a loose cannon with no checks in his power in the company.
Hes also a sociopath that thinks AI is more important than us
1
5
57
2,769
@pubity It just proves finance people are dumb, and just a bunch of inside traders, not able to make a cent without cheating.
They are low IQ and AI was just above the threshold for them to believe in it, unlike VR for example.
They are FucTX, and it's gonna be wild.
They are low IQ and AI was just above the threshold for them to believe in it, unlike VR for example.
They are FucTX, and it's gonna be wild.
1
0
16
1,951
@pubity Maybe they need to ask their AI how to operate at a minimum break even.
Did they think of that?
They probably didn't, did they? They just ask for more investors, right?
I wish I was so wealthy that I could afford to invest $1 billion and still be fine if it went up in flames.
Did they think of that?
They probably didn't, did they? They just ask for more investors, right?
I wish I was so wealthy that I could afford to invest $1 billion and still be fine if it went up in flames.
0
0
8
832
@pubity Build to solve real problems people care about first, the money will follow.
0
1
2
113
@pubity burning $278B in negative cash flow while aiming for $350B revenue is wild
0
0
3
627
@pubity Please let it fall, please let it fucking fall, Jesus Christ, my god, i beg you
0
0
3
401
@pubity Ever thought about making it more power resource efficient? instead of making it more powerful and trying to figure out how to gain profit from it keep the current power and price range and make it more efficient
0
0
3
973
@pubity Step 1: wait until 2035 or 2040 to start up ai, starting it up early was a terrible decision and cannot be profitable until the audience is fully comfortable with it and wants it
1
0
3
1,332
@pubity The only people talking good about all this ai bullshit are the ones selling it 馃ぁ elon musk, mark zuc, sam, etc etc
1
0
4
691
@pubity How much do you need to be in the hole to understand AI as we know it isn't working?
0
0
2
371
@pubity Meanwhile, AI MLs are getting so refined and the data needed for them will be so small and specific to the tasks the MLs are needed for.
tiny open source models will be the norm on our phones. Need a coding AI? Get this LLM and .txt files.
Data centres will be obsolete.
tiny open source models will be the norm on our phones. Need a coding AI? Get this LLM and .txt files.
Data centres will be obsolete.
0
0
3
553
@pubity At this point, even being profitable wouldnt save them. Time for them to call it quits and shut it down.
0
0
1
42
@pubityt hall's pad is still dirt. Typical MV switchgear already quotes about 52 to 80 weeks, and large transformers more often run multiple years. Some of that infrastructure is already getting ordered, long before the software covers the power bill. 2029 is when that spend has to start paying for itself.
0
0
1
742
@pubity If this bastard would be a decent person, he would shut down the AI shit right now.
But he is a rich bastard that sexually abused his sister, so he rather sees everything burn down.
But he is a rich bastard that sexually abused his sister, so he rather sees everything burn down.
0
0
1
479
@pubity fuck being profitable! Take every rich investors money you can and put it towards something that benefits everybody for free, thats the real tax on rich.
the overall benefits is their return on investment, it doesn't always have to be money gained.
the overall benefits is their return on investment, it doesn't always have to be money gained.
0
0
0
10
@pubity Shut it down we could be investing in something with real benefit to society instead.
0
0
0
277
@pubity Reach a 5 hour limit in 10 minutes just by reading code at this point. At this time it鈥檚 better to just do local modules.
0
0
0
405
@pubity Sounds like a bet that inference demand outruns the debt curve, the way AWS carried Amazon through its capex years. My honest question is who plays the AWS-equivalent anchor tenant for OpenAI at that burn.
0
0
0
148
@pubity Because chat gpt is free, doesn't require sign up to use so many people use it in colleges/ usts systems too without sign up.
Also, millions of users are using free yearly subscriptions too and the paying users are mostly inclined towards claude specially enterprises.
Also, millions of users are using free yearly subscriptions too and the paying users are mostly inclined towards claude specially enterprises.
0
0
0
615
@pubity The economics get difficult when model capability keeps getting cheaper while the cost of building frontier infrastructure keeps rising. OpenAI has to turn that gap into enough revenue before the spending catches up.
0
0
0
612
@pubity I think step number 1 is "start a for-profit organization". It's common sense that making money with non-profits is harder.
1
0
0
1,219







































