<!-- wp:paragraph -->
<p>I have taught Business at a number of Community Colleges and the intro to Business classes always has one chapter on Economic concepts (generally the first chapter) and one on Finance concepts. However, we don't really do math. Instead students get taught very general stuff, like the types of economic systems, monopolistic and competitive aspects, the business cycle, productivity, GDP and supply and demand. <br>Very useful stuff, but none of the rest of the chapters ties back into it so I'm not sure what the students take away from their studies.</p>
<!-- /wp:paragraph -->
Show full content
I have taught Business at a number of Community Colleges and the intro to Business classes always has one chapter on Economic concepts (generally the first chapter) and one on Finance concepts. However, we don’t really do math. Instead students get taught very general stuff, like the types of economic systems, monopolistic and competitive aspects, the business cycle, productivity, GDP and supply and demand. Very useful stuff, but none of the rest of the chapters ties back into it so I’m not sure what the students take away from their studies.
<!-- wp:paragraph -->
I used to work for data centers, for example, for one of the major NYC banks with ATMs. Those are generally called "on prem" data centers, i.e., they are in buildings owned or leased by the banks. I don't have current statistics but I don't believe many banks or other financial institutions host their data "in the cloud", i.e., in hyperscaler data centers. All those on prem data centers have backups, e.g., in NJ or upstate NY, but even if consolidated would not equal the size of the proposed Stratos center.The current concern with "AI" data centers is with those hyperscaler data centers and particularly with those that are devoted exclusively to LLM processing or to maintaining surveillance data, rather than helping someone navigate daily traffic (unfortunately I suspect some of those hyperscalers do both kinds in the same data center).
<!-- /wp:paragraph -->
Show full content
I used to work for data centers, for example, for one of the major NYC banks with ATMs. Those are generally called “on prem” data centers, i.e., they are in buildings owned or leased by the banks. I don’t have current statistics but I don’t believe many banks or other financial institutions host their data “in the cloud”, i.e., in hyperscaler data centers. All those on prem data centers have backups, e.g., in NJ or upstate NY, but even if consolidated would not equal the size of the proposed Stratos center.The current concern with “AI” data centers is with those hyperscaler data centers and particularly with those that are devoted exclusively to LLM processing or to maintaining surveillance data, rather than helping someone navigate daily traffic (unfortunately I suspect some of those hyperscalers do both kinds in the same data center).
<!-- wp:paragraph -->
<p>As I was reading today's article on data centers, I noticed one of the related articles at the bottom was a previous article by Scott Buchanan on Dec 9, 2025 that had already covered Google's TPUs. A quote and link to the article.</p>
<!-- /wp:paragraph -->
<!-- wp:paragraph -->
<p>"What is going on? It seems that the market is buying the narrative that Google’s Tensor Processing Unit (TPU) chips are a competitive threat to Nvidia’s GPUs. Last week, we published a tutorial on the technical details here. Briefly, Google’s TPUs are hardwired to perform key AI calculations, whereas Nvidia’s GPUs are more general-purpose. For a range of AI processing, the TPUs are faster and much more energy-efficient than the GPUs."</p>
<!-- /wp:paragraph -->
<!-- wp:paragraph -->
<p><a href="https://economistwritingeveryday.com/2025/12/09/googles-tpu-chips-threaten-nvidias-dominance-in-ai-computing/" rel="nofollow ugc">https://economistwritingeveryday.com/2025/12/09/googles-tpu-chips-threaten-nvidias-dominance-in-ai-computing/</a></p>
<!-- /wp:paragraph -->
<!-- wp:paragraph -->
<p></p>
<!-- /wp:paragraph -->
Show full content
As I was reading today’s article on data centers, I noticed one of the related articles at the bottom was a previous article by Scott Buchanan on Dec 9, 2025 that had already covered Google’s TPUs. A quote and link to the article.
“What is going on? It seems that the market is buying the narrative that Google’s Tensor Processing Unit (TPU) chips are a competitive threat to Nvidia’s GPUs. Last week, we published a tutorial on the technical details here. Briefly, Google’s TPUs are hardwired to perform key AI calculations, whereas Nvidia’s GPUs are more general-purpose. For a range of AI processing, the TPUs are faster and much more energy-efficient than the GPUs.”
<!-- wp:paragraph -->
<p>Fascinating. I have not thought to worry about data centers dematerializing, but I have worried about the internet going down, which I think would amount to the same results. Hence, I keep my files and have as much of my software (Word, Excel, etc.) on my PC, not in the cloud. And (I know this is so quaint) make periodic backups of my file on an auxiliary physical storage unit.</p>
<!-- /wp:paragraph -->
<!-- wp:paragraph -->
<p>The situation in Iran, with internet shutdown, is a cautionary tale. I have a little old shortwave radio, left over from pre-internet era. On a good night I can pick up signals from thousands of miles away (they bounce off the ionosphere).</p>
<!-- /wp:paragraph -->
<!-- wp:paragraph -->
<p>I have fantasized about growing my own food in case of TSHTF, but have been too lazy. Once you've traded away your extra AA batteries, I wonder what folks would have to offer in trade for food and gasoline? Offering to run a marketing spreadsheet or an econ model for the proprietor probably would not cut it.</p>
<!-- /wp:paragraph -->
Show full content
Fascinating. I have not thought to worry about data centers dematerializing, but I have worried about the internet going down, which I think would amount to the same results. Hence, I keep my files and have as much of my software (Word, Excel, etc.) on my PC, not in the cloud. And (I know this is so quaint) make periodic backups of my file on an auxiliary physical storage unit.
The situation in Iran, with internet shutdown, is a cautionary tale. I have a little old shortwave radio, left over from pre-internet era. On a good night I can pick up signals from thousands of miles away (they bounce off the ionosphere).
I have fantasized about growing my own food in case of TSHTF, but have been too lazy. Once you’ve traded away your extra AA batteries, I wonder what folks would have to offer in trade for food and gasoline? Offering to run a marketing spreadsheet or an econ model for the proprietor probably would not cut it.
<!-- wp:paragraph -->
<p>I gave up on INTC back in the end of 2024, missed out on all this: <a href="https://economistwritingeveryday.com/2024/11/14/i-give-up-standard-poors-wins/" rel="ugc">https://economistwritingeveryday.com/2024/11/14/i-give-up-standard-poors-wins/</a><br><br>Still hold some MU though</p>
<!-- /wp:paragraph -->
<!-- wp:paragraph -->
<p>re: CPU vs. GPU vs. TPU: The Evolution of AI Processors</p>
<!-- /wp:paragraph -->
<!-- wp:paragraph -->
<p>Nvidia seems to have lucked out that their GPUs used in graphics cards would be useful first for Bitcoin mining then followed by training LLMs. However, a better chip for training LLMs is the Tensor Processing unit (TPU) made by Google. Here's a link to a good article:</p>
<!-- /wp:paragraph -->
<!-- wp:paragraph -->
<p><a href="https://medium.com/@tahirbalarabe2/what-are-google-tpus-tensor-processing-units-a07491712a2e" rel="nofollow ugc">https://medium.com/@tahirbalarabe2/what-are-google-tpus-tensor-processing-units-a07491712a2e</a></p>
<!-- /wp:paragraph -->
<!-- wp:paragraph -->
<p></p>
<!-- /wp:paragraph -->
Show full content
re: CPU vs. GPU vs. TPU: The Evolution of AI Processors
Nvidia seems to have lucked out that their GPUs used in graphics cards would be useful first for Bitcoin mining then followed by training LLMs. However, a better chip for training LLMs is the Tensor Processing unit (TPU) made by Google. Here’s a link to a good article:
In reply to <a href="https://economistwritingeveryday.com/2026/05/09/the-empire-strikes-back-against-ai-cheating/comment-page-1/#comment-10987">Scott Buchanan</a>.
Eye tracking is usually in the laptop webcam. Second camera is to monitor cheating materials in their environment.
<!-- wp:paragraph -->
<p>Could you elaborate on " The phone students were using to cheat are now being marshalled as a “second camera” for remote test proctoring " ? Do you install app on their phone that watches their eyes?</p>
<!-- /wp:paragraph -->
Show full content
Could you elaborate on ” The phone students were using to cheat are now being marshalled as a “second camera” for remote test proctoring ” ? Do you install app on their phone that watches their eyes?
<!-- wp:paragraph -->
<p>I use 50% acidity white vinegar, but only on pavement and pavers. It works on the same day and often gets to the roots. It all eventually evaporates, causing no damage to non-plants. Be careful. It’s super strong and can irritate humans. Also, you’ll find out fast if your pavement isn’t clean, because you’ll suddenly have very clean splotches. My friend got a letter from the HOA because they realized that his pavers were dirty and not just darker.</p>
<!-- /wp:paragraph -->
Show full content
I use 50% acidity white vinegar, but only on pavement and pavers. It works on the same day and often gets to the roots. It all eventually evaporates, causing no damage to non-plants. Be careful. It’s super strong and can irritate humans. Also, you’ll find out fast if your pavement isn’t clean, because you’ll suddenly have very clean splotches. My friend got a letter from the HOA because they realized that his pavers were dirty and not just darker.
<!-- wp:paragraph -->
<p>The research paper is closely associated with what we call academic jobs. Tyler might be expecting academic jobs to be killed, too. But if the jobs still exist then the papers are likely to continue. </p>
<!-- /wp:paragraph -->
Show full content
The research paper is closely associated with what we call academic jobs. Tyler might be expecting academic jobs to be killed, too. But if the jobs still exist then the papers are likely to continue.