The question above was asked in Robert Reich’s blog yesterday. The answer? Easy: ROI. An economic study showed that every dollar spent bribing politicians retuned an obscene return on that investment (ROI) in the form of tax advantages, business merger advantages, government regulation “adjustments,” etc. My memory is of $16 being returned on each $1 invested, but that is just from memory.
You’d have to be some kind of idiot to not invest in what pays off better than any other investment.
I remember when laws were passed to make selling an item below its cost to the business selling it was illegal. (Is it still?) This was called predatory pricing or dumping, dumping usually referred to foreign goods being sold, where predatory pricing was used when domestic goods were being sold. The idea was a firm sells products at less than cost of manufacture or purchase in order to drive out competitors and/or to increase market share, then prices are jacked up.
American history is littered with such practices, one prominent implementor of such practices in our history was John D. Rockefeller, widely considered the primary “robber baron” of the oil industry, having founded and led the Standard Oil Company in 1870 to create a massive monopoly. Through ruthless tactics, vertical integration, and secret railroad deals, he controlled over 90% of US refineries by the 1880s. Rockefeller’s company controlled nearly all oil refining and pipelines by the 1880s, often referred to as “The Octopus.” (Standard Oil was ordered to dissolve by the Supreme Court in 1911 due to antitrust violations.)
Rockefeller was known to build service stations right near existing service stations of competitors, undercut their prices for gas and oil and then when those stations went belly up, bought them for cents on the dollar and converted them to Standard Oil stations.
The first major attempt to regulate such actions was the Sherman Antitrust Act, but applying it to large companies was and is difficult. The Clayton Act was also passed with the aim of promote fair competition and prevent unfair business practices that could harm customers.
So, why am I bringing this up now? Good question, I am glad you asked.
A rude awakening is about to dawn on early adopters of AI instruments. Wowed by these tools, at how they seem to make a company’s workers more productive and efficient as more work gets done for the same money or the same work gets done for less, the actual cost is going to be applied soon. AI companies have been selling their services for free or very low cost and have been losing money hand over fist as they “develop their markets.” We are talking billion dollar losses across the nation as a whole. They cannot keep doing that and stay in business.
You have probably heard of the energy costs, and “data centers” being built to create that industry. So far, those costs were being paid by venture capitalists, who expect a return on their money. The Wall Street Journal reported that OpenAI was on track to lose approximately $5 billion in 2024 against $3.7 billion in revenue — with annualized costs running close to $9 billion including computing and staff.
But all the news “that is fit to print” apparent doesn’t extend to stories like these: “In early 2024, a mid-sized legal firm in Chicago laid off eleven paralegals. The partners were proud of the decision. They’d run the numbers carefully. AI was handling document review faster and more accurately than the paralegals had. The math was undeniable. Nine months later, the firm’s AI costs had tripled, the paralegals had found other jobs, and the two senior partners who remained were doing document review themselves at two in the morning — because they could no longer afford the AI or the staff to replace it.” (Source: “Prepare for the End of Cheap AI” by Victor Mong, 2/26/2026 on Medium.com)
AI costs aren’t regulated (or even monitored) and if you have laid off employees (the only way to pay for AI) and changed procedures to utilize the AI, and trained your workers to do their jobs using that AI, you could find yourself, like the law firm above in a world of hurt.
Oh, but competition will save us, right? Those companies can just switch to a different AI, one that is cheaper. The point is that none of them can afford to be cheaper and they are all different enough that switching would be difficult. (How many of you would welcome a completely different platform to post things like this on? Or want to do your taxes using a completely different software package? Or.…)
Welcome to the Brave New World where AI hype and predatory pricing are delivering you into an unknown world <theme from The Twilight Zone plays>.
Postscript For the mathematically challenged: if AI cuts your operating costs in half for some aspect of your business (after the implementation costs, etc.) and you are smiling, if AI costs triple (as they did above), you are now paying 50% more than you were under your old system and the odds of those costs going down are slim to none.
I read Diane Ravitch’s blog every day (and if you are an education buff, so should you) and in today’s blog reporting on the decision of the New Hampshire legislature to exempt voucher students from taking the same tests public school kids take, the following goad appeared:
“So-called “education reformers” are all in favor of standards, tests, and accountability. Such a strategy, they insist, drives higher test scores.”
The pro-testing folks have put forth testing, lots and lots of testing, as the only way to improve the public schools. Can you pot the flaw in their thinking? In ordinary discourse it is usually stated in the form of a mis-attributed quotation: “Doing the same thing over and over and expecting different results is a definition of insanity.”
The problem is this: the fourth graders tested this year are a different batch of kids from those tested last year. You can compare the gross results, but not the individual results. Fans of educational research are familiar with the “pre-test v. post-test” research model. You start out testing kids as to how well they, say, jump rope. Then you “teach” them things that are supposed to increase their ability to jump rope, and then you test them again, in the “post-test.” Did the kids improve or not? You can tell this way. In these grand testing schemes, there are no pre-tests, only post-tests … on different groups of students.
Consider an industrial example. There is a team of guts stamping automobile fenders/ They place a sheet of flat steel into the maw of a massive hydraulic press and … Bam! The press stamps out a fender, which will need trimming to remove the excess metal, but that is not the job of the team. So, there is a stack of sheets. A team is timed as to how long it takes to stamp out 20 fenders. Then any errors are subtracted. Then the team is tested again … and again … and what do you expect to find? I expect consistency. The team know what they are doing. It is a repetitive task. They have been doing it for a long time and they have developed a recognizable pace of work.
But what if you wanted to improve the process? What would you do? To change the process at all you would need to change the starting material, the stacked sheets of sheet metal, the workers, or maybe the speed of the stamping press. If you don’t make any substantive changes in the materials or the equipment or the process, what should you expect? You should expect consistency.
This is what we want from fourth grade teachers, to elevate the third graders provided them to being fourth graders, ready to start the fifth grade. But what if the latest batch of kids come to school dead tired, hungry, and suffering from various maladies: toothaches, headaches, nausea, allergies, colds, flu, measles, etc. Far worse than the previous batch, so how does that effect results?
The error made by these “eduformers,” these politicians with their heads up their asses, is they think students are distinct and separated from their environment, when actually they are inseparable from their environment. If their parents are worried about being deported, or losing their jobs, or losing benefits, the kids will pick up that vibe and be less likely to be able to concentrate on school work.
If these politicians really wanted to help, they do not, they would address things like hunger, gangs in the neighborhoods, drug dealing near schools, etc. Just as if you let the temperature in the fender stamping shop descend to –42° you wouldn’t expect your workers to be as nimble, environments do count. This is why even greedy-ass corporations create positive work environments to keep worker productivity up. (At least until workers can be replaced by “AIs” or robots whose work environment requirements are minimal.)
In a recent Medium.com post, What Would Happen If a State Banned Insurance Companies?, not labeled “Members Only,” the following question was asked:
This article explores a single provocation: what would happen if a state stopped allowing private companies to stand between people and their medical care? We are talking specifically about medical insurance, not property, auto, life, or any other line. We are talking about the companies that decide whether a doctor’s orders get paid for.
In support of this question, the following was stated:
“The federal government already runs the largest health insurance program in the country. Medicare spent roughly 1.4 percent of its total expenditures on administration in 2016, covering salaries, fraud control, and patient outreach.² Private insurers, by contrast, spend approximately 17 percent of expenditures on overhead, a category that includes marketing, executive compensation, shareholder returns, lobbying, and the elaborate claims-processing infrastructure built to minimize payouts.”
“Between 2022 and 2023, care denials rose an average of 20.2 percent for commercial insurance claims and 55.7 percent for Medicare Advantage claims, driven in part by machine learning tools that automatically reject claims without clinical review.¹ Each of those denials followed the same sequence: a person paid their premium, a doctor ordered care, and the insurance company decided not to pay for it. Somewhere in a corporate office, that decision contributed to a quarterly earnings target. The system worked exactly as designed; the company made money by not doing the thing it charged the patient to do.”
It seems that the only barrier to insurance reform (or elimination) is the lining of politician’s pockets by these insurance companies. They have accumulated the funds to do just that using their current business model, no?
“Insurance companies perform exactly one function that matters: pooling risk. A large group of people pays into a common fund, and when someone gets sick, the fund covers the cost. The concept predates the modern insurance industry by centuries; medieval guilds pooled risk, and so did fraternal organizations. The underlying math requires nothing more than a large enough group of people contributing to a shared pot of money, managed by actuaries who calculate how much goes in and how much comes out. Actuaries work for governments, pension funds, and universities, and they do not need a publicly traded corporation to practice their profession.”
“ … and they do not need a publicly traded corporation to practice their profession.” Indeed!
The so-called AI Revolution has been promoted as a mechanism to make modern businesses more productive, but after many, many billions of dollars invested, recent studies have shown that AIs do not make business more productive, they in fact tend to make them less productive.
“How could that possibly be?” exclaimed all of the AI Hype Swallowers. It can be because of what is now being referred to as AI Hallucinations, which is a massive misnomer. What they describing is the tendency of the popular AI platforms to make shit up out of whole cloth for reasons unknown to anybody. These mistakes made by the AIs can be found and corrected, but the cost of doing so is prohibitive, this is because the “mistakes” make no sense.
AIs were supposed to be the next generation of computer code authors … except they make really bad whoppers of mistakes that human coders cannot easily find because no one in their right mind would make mistakes that bad. Think of a car mechanic who is diagnosing a problem you are having with your car. If it is not running, is the engine getting fuel? Is the engine locked up? Are the cylinders showing the same compression? Is there spark getting to the cylinders? It wasn’t rocket science and many, many mechanics became adept at diagnosing problems. Then imagine an alien landing their flying saucer taking it in for repairs because the faster-than-light drive seems not to be working. What should the mechanic do?
Currently the cat was let out of the bag, in Forbes Magazine, no less:
“Tech companies cannot profit from their AI investments without replacing human workers.” (Geoffrey Hinton, who literally invented the neural networks that power modern AI)
So the business case for AI isn’t “AI will help workers be more productive.” It is “AIs will replace workers entirely, and we’ll pocket the salary savings.” That the AIs will not be able to do those jobs as well is irrelevant, profits are profits.
And for those thinking that the truth will set you free, please consider the wisdom of the great Cory Doctorow:
“AI cannot do your job, but an AI salesman can 100 percent convince your boss to fire you and replace you with an AI that can’t do your job.”
I mean what the fuck to the “bosses” know about what really makes their business work? All they seem to know is “workers expensive, workers bad!” Well, that and “Vote Republican.”
And why use AI to replace 100 dollar-a-day workers when they could be used to replace CEOs, CIOs, Presidents and Vice-presidents, Board Chairmen, etc. AIs are already adept to slinging bullshit memos full of meaningless jargon, so … well, we are talking efficiency here!
According to a post by Will Lockett on Medium.com:
A recent MIT study found that 95% of AI pilot projects fail to yield meaningful results. For the 5% where AI yielded results, the improvements were marginal, and the AI was used to augment highly constrained back-office jobs, not replace workers as promised. A METR report found that AI coding tools, which are meant to be the most promising application for generative AI, actually slow developers down. The inaccuracy of these models means they repeatedly make very bizarre coding bugs that are highly arduous to find and correct. As such, it is quicker and cheaper to get a developer to code it themselves. Research has even found that for 77% of workers, AI has increased their workload and not their productivity, leading to burnout and decreased performance.
Add to that billions of dollars have been spent (in the US, China has spent less to greater effect, way less) developing the latest, “new, and improved” AI programs and it has become clear that their performance has plateaued (law of diminishing returns?).
The promises made for AI have been highly speculative and, who knows, they may eventually come about, but the large language/generative models currently in vogue have tapped out, without answering important questions such as why do AIs make up shit?
The danger comes not from AIs taking your job, but from the AI bubble, investors have pumped so much money into non-performing companies that a gigantic stock market “bubble” has formed, one that can bring the entire economy down when it bursts. If I may quote Mr. Lockett again “You see, AI critics like myself have been comparing the AI bubble to the infamous dot-com bubble of the late ’90s, but we have been sorely wrong. It is far, far worse. The dot-com crash will seem like a picnic in the park in comparison to what is to come.”
Once again, Big Tech is going to crash our economy, and like the last time, no one will go to jail (I think one bank president did from the savings and loan bubble, but none of the others, including the egregious real estate cheats from 2008.
Hang onto your wallets and purses, ladies and gentlemen, this ain’t gonna be small potatoes.
(This is a member-only story, but if you are a member of Medium.com, here is a link: AI Will Destroy Everything, But not in the way you think it will.)
Uncommon Sense 
Trump Not Smart Enough to Be Br’er Rabbit
Tags: corruption, hypocrisy, politics, Republicans, tax the rich
As reported in The Guardian: “… at his more than hour-long pre-cabinet press conference on Thursday morning, Trump denied that the US was ensnared. He reiterated that the military campaign was well ahead of schedule. The Iranians know they have a disaster on their hands, he said, adding that ‘they were begging to negotiate, not me’. He said: ‘If they don’t negotiate, we are their worst nightmare. I am the opposite of being desperate’.”
He’s lying.
When Trump re-named the Department of Defense,
illegally of course, as the Department of War,
he wasn’t kidding.
This Briar Patch was not one that Trump asked to be thrown into when cornered. He jumped in himself! (Br’er Rabbit never would have!)
When Trump renamed the Department of Defense, illegally of course, as the Department of War, he wasn’t kidding. I do notice that those edits of those titles (especially Secretary of War) are not being used much now, because Trump is insisting that this is not a war he has started.
The tools Trump is used to using: lying, suing, ignoring, etc. are not working. The whole world is watching (and learning how to defeat the U.S. in future engagements), and calling out Trump’s lies (see above), and he cannot ignore what is happening because it is of no consequence when it is clear his war is affecting Republicans’s chances in the midterm elections coming this fall. Trump thought he had captured a golden idol, only to find out that he had a tar baby on his hands.