AI: How to Win Big Every Time
March 2, 2026
Another dinobaby post. No AI unless it is an image. This dinobaby is not Grandma Moses, just Grandpa Arnold.
I read “Amazon & The Cost of (AI) Lateness.” When I saw the title in my newsfeed, my dinobaby mind translated it to mean, “Here’s THE justification for spending money on “if we build it, they will come” concepts and a fervent (almost fanatic belief) in the “next big thing.” But that’s just my view.

Thanks, Venice.ai. Good enough. Do you like pink?
The write up says:
Amazon, on the other hand, is paying a big price for being slow, lumbering, bureaucratic, and woefully slow in getting on the “AI” bandwagon. The Seattle giant reminds me of a rich middle-aged guy from the hinterland who has finally discovered Visvim. The right way to look at today’s news is to see what each dollar actually buys to determine who showed up early versus who showed up dumb and rich.
For those not in the know, “Visvim” is Japanese clothing now “up to 90 percent off.” I am not sure if that metaphor means those early AI billions are buying more, or the outfits that enter the AI hype space can get a discount. My hunch is that the idea is that Microsoft’s jumping into AI is really a super investment. Yeah, I assume although some Copilot users might wonder why there is AI in Notepad. Plus, there is the baked in “make up information” flaw that goes with the Google “All You Need Is Attention” approach to information.
The write up states:
Last quarter, Microsoft’s net income grew by $7.6 billion from its OpenAI investment alone. The company also has $625 billion in contracted future obligations, nearly half of which comes from OpenAI. And he loses nothing with the new deal the two companies jointly announced. Microsoft still has the option to participate in this round too. I mean, the whole $110 billion isn’t done.
Is there any accounting magic going on in the Microsoft numbers? Accountants are paid to be magicians. It has been and will continue to be easy to fool people with mathy stuff. Aren’t those Redmond wizards doing some re-thinking about its approach to smart software? Nah. AI is organized at Microsoft, the code for updates is better than ever, and the users like smart software… everywhere.
What’s Amazon getting for the money it is pumping into AI in general and OpenAI specifically? Here’s the answer:
Amazon is really getting bupkis.
Maybe but it did get written up as a loser laggard in “Amazon & The Cost of (AI) Lateness.” That is not bupkis. A “bupkis” to some means “nothing at all.” To others, “bupkis” is the name of a failed Peacock comedy.
The write up makes clear that pumping money into AI years ago is the key to success and blocking the laggards. Does that mean that outfits offering access to excess AI compute are going to struggle for revenue? That seems dicey: Betting that AI compute surplus will be the killer business. Does it mean that misfiring Apple is just fooling itself with its alleged Google AI love fest?
The write up concludes:
There is no greatness in lateness.
I would point out that before Google found inspiration in the GoTo, Overture, Yahoo ad approach, Google was indeed late to the online advertising game. What happened? An alleged monopoly and now a Code Red thrust into one of the top spots in the AI game. Yeah, late. That’s for losers.
Observations:
- AI is likely to be one of those winner take all online games. The folks are spending hard because that’s their competitive advantage. Today’s AI runs on essentially similar methods.
- AI outputs mistakes and these occur no matter what because the underlying system depends on probabilities. A verified datum is different from and AI output a certain percentage of the time.
- Today’s AI is not the end of making zeros and ones act like a human. The journey leads to the same fast talking around problems. A new idea, not a replication of the Transformer approach, seems to be necessary in my opinion.
Net net: This essay justifies big tech spending since Microsoft’s 2022 announcement. The spending is notable as is the assumption “if we build it, they will come.” Yep, sure.
Stephen E Arnold, March 2, 2026
Comments
Got something to say?

