Subject: A thought about AI
WendyBG has been commenting on the increasing amount of debt (much of which hidden due to accounting methods used) starting to be assumed by the obvious companies who ware funding various AI efforts. Her concern is that this debt may be the catalyst which foments the next stock market crash. Plausible, but that's not the point of this post.

Like many of you, I'm old enough to remember the fallout in the personal computer venders during that revolutionary technology shift. For every winner, there were dozens (hundreds?) of firms which dropped during the 1980's into the distant fogs of memory. The same thing happened again, between 2000 and 2002 when, networking/internet first became the apple of investor's eyes and then crashed with the vast majority of the firms tied to that wonderful (truly) opportunity who crashed and burned.

The companies who pick the right AI horse stand to make an enormous amount of money someday. In the meantime, their profitability will be zip as their cash (and borrowings) are poured into AI ventures. The ones who back the losing horses, will have not generated real returns to their stockholders for years and possibly risked pissing away the whole shebang.

It just struck me that Nvidia, like Taiwan Semiconductor, are selling to the guys doing the build-out. That means that they are on the side of the equation receiving some of the money the others are spending. While the software guys are going into hock, but unless the chip manufactures are drinking the Kool-Ade and pouring their profits into the same pit, they stand to benefit regardless of who wins. Their major risks (excluding future competition from China and elsewhere) is, after the crash, receivable issues with bankrupted companies and competing against "discovered" overstock of their chips being liquidated by others.

Other winners (with the same potential receivables risk) will be the firms supplying the goods and services required to build the required data centers as well as electrical utilities which can back-charge their distributed customer base to finance the larger capacity required by the electrical grid. (The alternative being plant-scaled nuclear power plants attached to those data centers.

Looking into the clouded crystal ball, until/unless AI becomes socialized, the changes it will make in the economic structure of our population could be cataclysmic. Cutting through self-delusional nonsense, similar to the fact that few children today can perform simple math procedures without a calculator, the critical thinking that traditional educational methods have fostered for a very long time, children using AI to "assist" them with their schooling will have a very different set of personal capabilities, (excluding their AI "assistant") than could be taken for granted in the past.

I find myself wishing for AI while I listen to on-hold music and the encouragement of "your call is important to us" for 20 minutes and then being told that a supervisor will call be back sometime over the next 5-7 business day. I find myself wishing for a car that takes me where I want to go without my intervention. I want a robot that will do household chores, an automatic chef, pick the most efficacious air flights, a way to do hundreds of tasks - large and small - that others provide for me today. Each of those will cost "someone" their job. Sure building more AI capacity will require humans, but I suspect the tradeoff will be losses of jobs of professionals such as doctors, lawyers, engineers, CPA, etc. In short, our society post-AI revolution could be significantly different (as I sit here typing on a PC attached to a network, while chatting on my mobile phone with a wristwatch attached - all of which were science fiction topics in my youth).

As the late, great, Yogi Berra once (purportedly) said "It's tough to make predictions, especially about the future."

Jeff