Smart Software Fix: Cash, Lots and Lots of Cash
August 19, 2025
No AI. Just a dinobaby working the old-fashioned way. But I asked ChatGPT one question. Believe it or not.
If you have some spare money, Sam Altman aka Sam AI-Man wants to talk with you. It is well past two years since OpenAI forced the 20 year old Google to go back to the high school lab. Now OpenAI is dealing with the reviews of ChatGPT 5. The big news in my opinion is that quite a few people are annoyed with the new smart software from the money burning Bessemer furnace at 3180 18th Street in San Francisco. (I have heard that a satellite equipped with an infra red camera gets a snazzy image of the heat generated from the immolation of cash. There are also tiny red dots scattered around the SF financial district. Those, I believe, are the burning brain cells of the folks who have forked over dough to participate in Sam AI-Man’s next big thing.
“As People Ridicule GPT-5, Sam Altman Says OpenAI Will Need ‘Trillions’ in Infrastructure” addresses the need for cash. The write up says:
Whether AI is a bubble or not, Altman still wants to spend a certifiably insane amount of money building out his company’s AI infrastructure. “You should expect OpenAI to spend trillions of dollars on data center construction in the not very distant future,” Altman told reporters.
Trillions is a general figure that most people cannot relate to everyday life. Years ago when I was an indentured servant at a consulting firm, I worked on a project that sought to figure out what types of decisions Boards of Directors of Fortune 1000 companies consumed the most time. The results surprised me then and still do.
Boards of directors spent the most time discussing relatively modest-scale projects; for example, expanding a parking lot or developing of list of companies for potential joint ventures. Really big deals like spending large sums to acquire a company were often handled in swift, decisive votes.
Why?
Boards of directors, like most average people, cannot relate to massive numbers. It is easier to think in terms of a couple hundred thousand dollars to lease a parking lot than borrow billions and buy a giant allegedly synergistic company.
When Mr. Altman uses the word “trillions,” I think he is unable to conceptualize the amount of money represented in his casual “you should expect OpenAI to spend trillions…”
Several observations:
- AI is useful in certain use cases. Will AI return the type of payoff that Google’s charge ‘em every which way from Sunday for advertising model does?
- AI appears to produce incorrect outputs. I liked the application for oncology docs who reported losing diagnostic skills when relying on AI assistants.
- AI creates negative mental health effects. One old person, younger than I, believed a chat bot cared for him. On the way to meet his digital friend, he flopped over dead. Anticipative anxiety or a use case for AI sparking nutso behavior?
What’s a trillion look like? Answer: 1,000,000,000,000.
How many railroad boxcars would it take to move $1 trillion from a collection point like Denver, Colorado, to downtown San Francisco? Answer from ChatGPT: you would need 10,000 standard railroad boxcars. This calculation is based on the weight and volume of the bills, as well as the carrying capacity of a typical 50-foot boxcar. The train would stretch over 113.6 miles—about the distance from New York City to Philadelphia!
Let’s talk about expanding the parking lot.
Stephen E Arnold, August 19, 2025
Comments
Got something to say?
-
- Subscribe to Beyond Search
Feature archive
News archive
-
