One more time… typo not type. I know it is late and I am tired. But something seems off with auto-correct (and me). Anyway, one more time with feeling:
Here is a reduced-typo version (I’m not looking at this anymore. There are probably still typos. I will do better on my next new post – I hope.):
First, thanks to Ed A. for bringing this article to my attention. I lived in Silicon Valley and worked in the semiconductor industry during the dot-com and internet bubble. I remember how fiber optic companies were all the rage. Then the bubble burst and there was frequent talk of all the “dark fiber” in the ground. Dark fiber are all the fiber optic cables laid in the ground for the “guaranteed” growth in internet traffic. Well, like all technologies, adoption takes time.
The WSJ article provides other examples of hype-driven overspending and subsequent bubble bursting such as UK railroads. It did not mention another recent and prominent example: 3D printers. If you forgot about the hype and subsequent bubble burst of 3D printers, all you need to do is look at a 5 or 10 year chart of DDD or SSYS to jog your memory.
So in short, an unprecedented amount of growth is priced in “AI” stocks and data center construction projects. There’s also a lot of potential damage to the rural areas where these centers are going up: higher electric bills, water pollution, air pollution, increased housing costs, etc.
If you read the book AI Snake Oil, it too talks about not only the hype but the drivers of the hype: the very companies selling you their AI products, the media that needs clicks and views, and public figures incorrectly placed on pedestals as knowing the future.
Have a look at the WSJ article. The numbers are staggering. Then put it in the context of the circular financing we saw announced over the past week or so (e.g., Nvidia invests $100B in Open AI so Open AI can buy $100B of Nvidia chips, Open AI striking a “deal” with Oracle for $300B for “planned” purchases). Lots of plans and intents. Lots of data centers being constructed with no customers too.
Maybe “this time is different.”
Maybe it is not.
-Dr. Moore
https://www.wsj.com/tech/ai/ai-bubble-building-spree-55ee6128
Efficient Minds™ 
Anyone who’s tried to use A.I. for anything further than a novelty has quickly seen its limitations, limitations that have stayed unbroken for at least an entire year, which is an aeon in the tech sector.
Now we have server farms paying utility bills to essentially construct different amalgamations of existing data with no true originality or critical thought. Essentially nothing new is being produced at great cost to someone somewhere and the technology is not advancing to where it will be able to produce before the bubble bursts.
AI has to get to a point where it can produce something humans want to consume or the lack of revenue generating applications will soon hit the limit of what the optimistic investing has provided for.
You can only butter so much bread on replacing telemarketers customer service representatives with chatbots.
Thank. you Gareth for the comment. I told students today that after posting that blog (3 times due to typos) I have since turned off autocorrect and auto-punctuate on my phone and laptop. I told students this deemphasis on “AI” and “AI”-like features leads to more mindfulness of what one writes. I even came up with a hashtag for this concept / movement: #De-AI
I dared students, and anyone reading this, to turnoff autocorrect and auto-punctuate for a week and see how more engaged one becomes with their writing. I think that is good for the brain. And I am not evening talking about IBM’s article “When AI thinks for us, the brain gets quieter”: HTML.