Lindsey Graham funeral shows that most corrupted and reviled politicians in America will get grand State funerals of a Hero. America is a bankrupt country.
The oligarchs including Donald Trump control the Republican Party. Elon Musk, Peter Thiel, Jeff Bezos screw their workers and continue to block all legislation that will make housing, food, energy, health care, etc. more affordable. They are all cruel and hateful greedy bastards.
What I can't understand is why, as billionaires, they want more money. If you have a billion and spend at a rate of $5000 a day and assuming zero investment returns on your money, it takes around 540 years to run through it. Dunno about you but I could get along fine on $5000 a day.
They want the power that accruing more extreme wealth brings.
Harv Eker referenced the psychology of people who come into extreme wealth. For a few years, they typically go wild, buying expensive cars, clothes, planes, houses, etc.
But then they get bored, and look for a reboot. The psychologically healthy ones look for ways to instead help the country, the world. The twisted ones instead push for more wealth, more power.
The thread below has been hijacked, so regarding Dr. Krugman''s post: Financial markets have a mind of their own having nothing to do with us little people. Big money needs places to go and it's AI.
The stock market climbs a wall of worry and predictions of a crash seem greatly exaggerated. If it happens, that might have more to do with the stupid, metastacizing and endless war in the Gulf.
Just saw an apt line attributed to David Cochrane, a Canadian writer, in a comment following Jen Rubin's incisive Substack today: "America is not like this because Donald Trump is president. Donald Trump is president because America is like this."
I remember when I realized this, a while ago, and it was profoundly depressing. An election might help, but will not not solve what is wrong with America. Because it’s Americans themselves that are what’s wrong and that is not going to change in my lifetime.
Thank you. I remember that quite well, and the solidarity after 9/11. That was all pre-Covid. I think the pandemic broke America, because that served Trump well. Even by their very nature, pandemics are divisive and breed distrust; I remember David Brooks saying that right at the beginning of Covid, because it surprised me. Now it makes sense.
The people in "Miracle on the Hudson" and Trump enthusiasts could very well be the same, confused people. I do so hope that this confusion is temporary.
I think it also tells us about our schools! My parents could think logically, with basic training of how to apply their noodles. All of that type of education seems lost in recent decades. We must tune-up our children's education with common sense, civics and respect for one-another.
This reminds me of my parents who first registered as Republicans in the early 1950s, before the party embraced the Southern Strategy of racism and religious extremism. By 1980 they were fed up and stopped voting for Republican presidential candidates, but remained registered as Republicans for the primaries in an explicit effort to hold the line against the crazies. It actually worked for a while in my state, but the nationalization of politics eventually won out.
I remember my parents, when moving to Florida in 1953 from Maryland being nonplussed at not being able to register to vote as Republicans..... How far we have come! (I'm not saying which direction.........)
That idea isn’t going to get you anywhere. It is too academic and doesn’t help anyone except maybe the senators.
The Republicans aren’t the problem. They are obviously broken, but it isn’t the core issue. Nobody is speaking up for the working man. Trump is the closest thing, but he isn’t going to lift a finger for them. He will stiff them on the bill like he always does. I’m all for the Republican Party becoming that progressive front, but I’ll believe it when I see it. The DNC is more economically right wing than the Republicans. They are fighting even harder to avoid it. The working people are mad, and until someone is doing something for them, they are staying mad, and we are going to be in this hellscape indefinitely.
I was particularly struck by Vance's eulogy, where he recalled that he had approached Graham to vote for a particular project, thinking that Graham would lie, like any politician, but Graham told Vance he had already voted for it. Vance was surprised and told Graham he didn't know Graham was interested. "Oh I'm not," came the reply. "I did it because YOU wanted it." Vance was hugely impressed, he apparently thought this spoke to Graham's integrity, whereas it was merely the act of a weak man kissing the ass of a narcissist.
I haven't looked at the morning paper yet; did Lindsey get a big sendoff? Weird guy. Weird country. It was weird in different ways years ago, but 21st century America is . . . words fail me.
He was definitely sincere in his support of Ukraine. He took the train to Kyiv at least eight times to discuss matters and show his support. He pushed legislation to put more sanctions on Russia.
Graham was actually among the biggest lobbyist for US defense industrial complex - Trillion dollar hydra. Anywhere where US can market its merchandise, Lindsey was always supportive of that - He was a merchant of death.
And his alignment with Ukraine was just coincidental with his lobbying interests.
If US could sell its defense merchandise to Russia, Graham would've had 'sincere' support for Russia.
There was nothing 'sincere' about Graham. A sociopathic liar and a scoundrel of the highest order who deserves 7th circle of hell.
NSAlito….His “bads,” outweigh his “goods” by a long shot. He gave us the worst Supreme Court imaginable. He is a lying hypocrite. That is the bottom line.
It's just that I only learned of his special support for Ukraine the day before he died, and I want to remind my brain that warmongering sniveling toadies aren't necessarily ·all· bad.
I certainly wouldn't like the grand State funerals of a Hero that these corrupted and reviled politicians would get. It actually sickens me, but at least they'd be dead, good riddence.
Politicians who are veterans always have elaborate funerals. It doesn't hold much weight with everyone else. Just Graham's history in the Air Force increased his standing--or lying in state.
I recently saw the episode of Downton Abby where Lord Grantham bet on the wrong railroad & lost his wife’s fortune (whom he truly loved, but it helped that she had money when they married to prop up the estate's failing finances). The railroads were a revolution, but not all were successful, and it was hard to tell in real time how it would shake out. Lesson for today.
Important to note that Lord Grantham made the investment against the advice of his financial advisors. I believe they recommended more diversification.
I love that you use the term, "huevos". That's been one of my favorites for a long time. Is this commonly accepted forever? One of my first Spanish expressions was, "Dos huevos." I just commented about Andy Beshear and his missing "huevos"...
We're probably still a long way from weighing the benefits of technologies against the real costs of using them. It just never seems to be part of the equation to say, "this is really good stuff but we need to find ways of using it that minimize its impact."
This isn't even 'really good stuff" - except for the entertainment value of watching lawyers get sanctioned for using AI in their legal briefs which hallucinate case authorities/quotations. (The med mal cases that have or may result are less amusing...)
There is some really good stuff in niche areas but technology is rarely siloed limiting its use to things that have really favorable cost benefits. Actual environmental costs are not often considered.
This reflects what so many pundits either ignore or don’t know: most of us have already listened to Republican-MAGA relatives during holidays and family reunions, and we’ve listened to similarly minded classmates and coworkers. Listening more isn’t going to solve the fundamental disagreement. We’ve already listened and concluded we want different outcomes and facts matter.
I’ve read enough science fiction to know this AI thing can be very dangerous. And then there’s too much information about our own personal thoughts out there for the wrong people to know about.
While our domesticated brains are smaller, we have been able to use them more effectively by equipping them with more expressive language, education, and offloading cognitive load with technology like books (facilitated by printing) and computers. Our phenotypes have been really extended.
Back in the day, you had to be fairly sharp as well as attentive to multiply numbers written in the Roman numeral system. And it took a fairly long time to do so even if you could. These days positional notation, usually in base 10 make it a snap.
It is much easier to handle numbers with the Arabic script. Interestingly, my British handling of numbers for addition and subtraction is slightly different from teh way my son was taught in California.
Calculators are a huge boon to rapid calculations, and spreadsheets and computer programs vastly increase that facility. Yet when I taught university freshman math classes, I found that their reliance on calculators created what I called "number blindness". The students couldn't detect when they had made a mistake in a calculation. They had also lost the skills that I was taught in the 'slide rule era" to determine expected decimal places, create estimates of results using nearest 10s, and factor out large sets of numbers in divisions or equalities. All these techniques make handling some calculations easier, and ensure that the calculator answer is within th expected range.
So the calculator for contemporary students has meant that the cognitive effort that is now offloaded is also lost in the brain. This is the same as our ability to memorize prose and poetry has been largely offloaded into books.
This is something that makes the end of Bradbury's "Fahrenheit 451" so jarring. People preserving books by memorizing them word for word. As we know from the oral period before printing, this can not be done without error over generations, so that stories change over time. (It is like the Chinese telegraph game).
But our danger when putting texts into digital media, especially on cloud servers, is that it is prone to loss and deliberate manipulation. This is Orwell's warning in "1984" and the aims of the Party's Ministry of Truth. We can see the same desire in Trump's regime of trying to change history by removing historical exhibitions and removing or changing labels.
Do we really know that absolute brain size is the deciding factor in the success of a hominim species? Neanderthals were exquisitely adapted as hunter gatherers and their brain size no doubt reflected that. Somewhere along the way, sapiens evolved brains with the capacity, among other abilities, to solve differential equations. Maybe (and I’m just speculating) that brain didn’t need to be bigger, just wired differently.
I forget the exact numbers - but it goes something like this: A single human brain can perform (a bajillion) calculations per second and only needs to consume about 1,000 calories of energy (a Big Mac, side of fries an maybe a couple beers.) An equivalent AI Data Center needs (umpteen thousand) gallons of water and about (some absurd number) of kilowatt hours to perform the equivalent number of gigaflops. My point: AI will be great for some things, but it's not going to recreate, all by itself, the Beatles Catalog. Let's leverage the strengths it offers but not worry too much about the World becoming a scene from Terminator.
My husband has two engineering degrees&a masters in information systems.
The human brain has gotten smaller, our intellect has diminished&I don’t see any existing educational system doing anything to counter that bleak fact.
I’m aware of the capacity of the human brain.
I made a half humorous comment that you felt compelled to counter.
Don’t assume you’re winning arguments with me by attempting to school me.
I have a Commodore computer in the attic. Information systems evolve&exponentially are evolving.
I wish as in the “Dune” reference humans would make use of what we have.
Current GenAI may be able to reproduce Beatles catalogue, but it does not create anything on its own without a human prompting it. Not even a single word. GenAI systems have no thoughts, desires, plans nothing of that kind. Machines capable of that are still very much just sci-fi and I hope they will remain sci-fi in the foreseeable future (ideally forever).
@Martin. The systems imitate rather well. I recently asked Perplexity whether a certain garment made me look fat. In return I got a whole sermon about how I should not body-shame someone.
Again, my point is that if you did not ask Perplexity the question, it would’ve never said anything. The LLMs it uses have no intentions, desires, plans, opinions on their own. They can give expansive answers addressing aspects you did not explicitly ask about. However, you must ask (or at least say) something. Otherwise they stay quiet.
During a test, an OpenAI model hacked into Hugging Face’s computers when searching for the answer. That shows planning. It also demonstrates, if not a desire to do well on the test, something close enough to a desire to do well on the test that the distinction may not be of practical importance.
My point is that the current AI must be given an explicit goal to do anything. It does absolutely nothing on its own. Once given a goal it can do planning, trial and error testing etc.
Great episode of Downton and may very well be a cautionary tale for AI. We’ll see how all this shakes out. I would be curious about Warren Buffett’s initial take on this AI “revolution.” Thanks for your comment!
I don’t know. What do you mean? Taking ideas that are out there and turning out a product that’s more functional is innovation, no? That’s not the same as eating off of a dead man’s table.
Apple used to be more innovative but lately they are not really first movers.
Since Steve Jobs passed, they have become more popularizers of ideas that are out there and proven than they are risk takers on the uncertain frontier. And they have been very successful at this.
Great episode of The Story of Money today about how the construction of the Erie Canal was financed with municipal bonds. Municipal bonds really are fascinating. Not all municipal bond issues succeeded, but I was surprised to learn how big the municipal bond market is, and how we will need them to finance all the infrastructure renewal that needs to happen. Great angle on how they relate to local democracy.
The railroads are a great example of irrational exuberance.
I always feel the need to post this link to 'Agents of Chaos' in moments like this, because you are correct. AI, though, increases those risks because we're asking it to make decisions.
In this article, red team hackers set up a controled experiment in which AI agents with decision making authority are tricked into making bad decisions. One outcome in the logistics realm: they are able to shut down shipments of food to grocery stores.
We need to really absorb the damage we can do to our system by trusting AI too early.
Downton Abbey’s storyline opens with the Titanic! Perfect metaphor for our time and yes that show was a lot of fun. I often compared myself to the ‘waitstaff’ at my last ft job, where I prepared communication materials for the company elites.
The railroads had a tremendous impact on the country and society but, as you point out, not all of them were successful. The same thing happened with the auto industry and probably with just about all new inventions. The problem from the point of view of the investor is -- which company will be successful and when (over what time period). Which AI companies will be successful and when?
Many a mother of a US girl with "new" money, worked hard to catch an impoverished Englishman with a title for said daughter. Edith Wharton started to write a novel about this phemomenon: The Buccaneers. (EDIT: this reply was meant to be posted elsewhere, @BKW.)
I just read an interesting piece on the circularity of investment between chip makers and AI producers that likened it to the model Enron used, which eventually led to the collapse of Enron. These guys are propping up the value of each others’ companies and the suggestion is that if any piece starts to unravel, it’s going to unravel quickly and it will be bad.
Indeed, pkidd! PK’s conversation with the AI expert at MIT, Paul Sodrowski, also mentioned this. And they’re using complex financial instruments like CDOs to finance their arrangements. Quite right! What could go wrong?
Say, I missed that talk and am having a hard time finding it. I tried searching the name in PK’s substack, but nothing comes up. Do you recall when it was? Is the name correct?
Just read this, think I may have understood some of it, especially the later parts on bubbles, etc. Will go back later and re-read if I can. But thanks very much for the link! And thanks, PK’s!
I love your handle. Whenever you reply to my comments, the notice says, “noone really replied to your comment…” There’s also a “Someone” out there. Other possibilities: someone actually, noone ever, noone ever really, just one a-hole, I, my dog, my next door neighbor (lucky you!), etc. 😂
Crypto consumes real world energy to calculate a random number on one end of the block chain with millions of computers, which could be calculated at the other end of the block chain by one computer in under a microsecond. I expect nuclear bombs are worse, but it is down there with military munitions for negative ROI.
AI has its moments for some tasks. So do spreadsheets and word processors. One would like to think that spending on AI would be commensurate with office suites, not national economies.
I now work part time as a parish administrator. For 23 years I worked for EDS (based in Plano TX) and HP after it bought EDS. I had a front row seat to the ENRON scandal as the project manager of the team charged with pulling documents for the SEC investigators.
I can tell you definitely: "the circularity of investment" generates a lot of smoke and mirrors that makes it really hard to identify what AI is really going to do... and also... who is going to do it.
Here is what I have been able to discern about AI so far...
1) AI is not going to take your job and is not going to make your job any easier. It will, however, change your job... a lot.
Take church administration as an example. Twenty years ago, people attended church in person, read the printed paper bulletin, donated with cash and paper cheques. Today, we want our personal preferences met. Those who prefer in-person still attend in-person, but they are only about half the population. Some (generally elderly) still prefer cash and paper cheques. So to reach the same people in 2026 as church did in 2006, we have to use video, Zoom, blogs, newsletters, electronic payment,... in addition to... everything that went before. Good that there is AI to transform content. But we in the parish office are running very hard in order to stay in one place.
2) The Tech behemoths of today will not dominate AI five years from now.
The AI implementations that solve real problems, for real people, will make real profit, and will eventually dominate. You will not know the names of these companies unless they solve a problem YOU have, as most of the problems will be very, very boring... such as better way to manage accounts payable, or scheduling the shipping of thousands of widgets from thousands of suppliers so that they all show up, at the same time, at a piece of land you've bought, in time for your contractor to assemble your house.
Moreover, along with contractor's hammer and spanners, AI will just be one of the tools. You will see the delivery vehicles and the guys with hammers. You will not see the AI that scheduled their arrival on site.
"You will not know the names of these companies unless they solve a problem YOU have, as most of the problems will be very, very boring..."
----
Back during the tech boom of the 1990s, the advice to people who wanted to launch a tech startup was, "First ask the question: Is my product a vitamin or an aspirin?" This was an analogy to vending machines in the airport, which sold a packet of two OTC painkillers for $2, because when you need it, you need it. There is no market for vitamin pills in an airport.
You are correct to point out that the generation now graduating university has more mental health issues than their parents because they are more socially isolated. More zoom and social media won't help... enough. In-person church activities will help more. But they see in-person church as their grandma's activity, nothing to do with them.
If you take the statement above and paste it into NotebookLM, the AI will respond. Some of the web links and documents it finds will be useful to a church trying to develop strategies for church leaders over the age of 60 to reach people under the age of 30, such as your neighbor. That is an example of AI helping.
When I was an IT project manager I often had clients who thought they could solve their problem if they just threw more tech at it. I would tell them problems are personal: A problem stops SOMEONE from doing SOMETHING that they want to do. If you cannot identify who to the point you call them on the phone, and what in enough detail you can put a dollar value on being able to do it, more tech is not going to help.
This is the situation of AI today: Too many executives expect AI to solve their problems without them having to do any hard thinking.
Another vote for Ed Zitron. Been following him since I stumbled onto his work the first of the year. It could stand some organizational and copy editing (and fewer f-bombs), but it's thoroughly researched and a blunt, clear-eyed critique of the AI media-and-finance frenzy. If he's right, it's not going to be pretty, and it could resemble the dot-com and housing bubbles combined, overlaid by hundreds of billions of Enron-grade financial shenanigans.
Exactly. Back in the day there was 'Nicolas Cage is a ripe pomegranate' trope about classifiers, namely that for any input they were obliged to produce some sort of output, and usually drawn from a smallish set options at that. Hence the caution that a classifier built to assess the ripeness of a fruit simply does not have the option to output 'Nicolas Cage' even when the input is, in fact, an image of Nicolas Cage.
The trope still holds, no matter how much more complicated and powerful the machinery is today: By its very nature, the product currently being hyped as 'AI' can only interpolate; that same nature guarantees it never extrapolate outside the range of the data it was trained on.
Also, the bed cover is not a patchwork quilt - the pattern is printed/on the cover. No big deal when considering the impressive discussion of the cat's photo, but a similar mistake in, say, a legal document (where AI is expected to earn its keep) could be a deal breaker.
I use three AI apps for comparison and translations. I give them the same text in English and ask for a German translation. Then I ask them to translate the text to English. Not one of them comes back to the original, and all three give different results.
Once tried to get a company to translate an operations manual into Chinese (was working there) and then get a different company to translate back into English. I did not get permission - I suspect we would have had a hard time understanding anything from the 2nd translation
There are stories -- most of them apocryphal, no doubt[1] -- about early translation programs going from English to another language and back again so that, for example 'Out of sight, out of mind' came back as 'Invisible idiot'.
"English as She is Spoke" was first published in 1855 by a Brazilian who took a Portuguese-French phrase book and a French-English dictionary and created a Portuguese to English phrase book. English missionaries found it in use as a text book in schools where they went to teach. Today, you can get it on Amazon, but in the "humor" category.
Today in the NYT there is a description of efforts to make an AI do the work of an administrative employee. Didn't go well, but I suspect a new human employee might have problems too until they get the hang of a job.
Meanwhile, dealers in rare and antique books are reporting that our tech lords are buying up old books to be ripped apart and fed into the scanners for AI. These people respect nothing, least of all you and me and what we might find a desirable future. Mars anyone?
Anyone else read the interview with Zuckerberg in the NY Times? The vitriol directed at Zuckerberg in the reader comments was something else. And it was comment after comment after comment. I guess the tech bros have fallen down a few pegs. Cue the tiny violin music. https://www.nytimes.com/2026/07/28/technology/mark-zuckerberg-meta-ai.html
The dealers really aren't in a position to complain. Taking apart old books and selling the individual pages as "art" has been going on for at least 50 years. I still have a few pieces that my mother bought in Europe in the '70s. Most have certificates of authenticity.
Vernor Vinge's sci-fi novel, "Rainbows End" has a subplot of the protagonist upset that his university library was being fed into a "chipper" and the fragments digitized and reconstructed into digital books. (A bit like Venter's shotgun genomic reconstruction for texts.)
I too am appalled at the idea that old books are destroyed to make it easier to and cheaper to digitize them. OTOH, this has been done before for a different purpose. Back in the 1990s at a San Francisco Book Fair, one could buy individual illuminated pages extracted from old religious texts. I don't even like buying used books with writing in the margins and highlights.
Nathan Myhrvold, ex-technology chief at Microsoft, was well-known for buying very large numbers of books each year. Far more than he could read, I am sure, but we cannot know that he didn't read some.
I still prefer reading physical books, but I probably read far more online, like this blog.
The massive expenditure on "AI" infrastructure has got to be the greatest misallocation of capital in world history. And that's to say nothing about the opportunity cost. At these prices, we could have had a full upgrade of our power grid and electrical generating facilities, for example.
AT&T used to be the phone company. They were broken up and the resulting company left with the name went on to build several amazing businesses that they always seemed to start a little too early and were never able to be successful. The assets would get sold off and someone else would make a mint off of what they built.
I wonder if the frontier AI companies or maybe those dumping billions to not catch up to them will serve the same purpose. Will Oracle or Facebook or SpaceX be left with all of this capacity but models that don't measure up to the market? Will other companies be able to take their capital and build the AI equivalent of Google that turns the equivalent of AOL or Yahoo into a zombie while the upstart thrives?
That Trump wants to ban Chinese open-source AI models is indicative of the threat they offer to big AI Labs like OpenAI and Anthropic. I have just read that the US is going to ban Chinese humanoid robots that seem to be advanced and much less expensive than US ones.
I wonder where the money will come from for the millions of people unemployed because of AI, to live their best lives sipping margaritas by the pool, after a hard day's shopping? Do you think the tech bros are saving up for this? Who will pay for your house? Your car? Your child's education? Elon will feudalize you; you will be scrounging in the dirt for a slug to eat.
If we do the wealth distribution right then we can all retire at age 55 and spend our time doing more fun stuff. You know share the work and share the wealth.
After reading all of Burry’s analysis, and then hearing Sam Altman say, “We’re going to let our AI figure out how to pay for itself,” I’m feeling like the future ain’t too bright.
As a Korean, I feel deeply saddened and concerned about the vulnerability of the KOSPI and KOSDAQ.
Finance, by its very nature, inevitably involves risk-taking in the pursuit of returns. However, Korea’s capital markets are relatively small compared with major global financial markets, and the scale and speed of the recent market turbulence have been truly alarming, given their limited capacity to absorb such abrupt volatility. I remain optimistic about advances in AI technology. From the perspective of financial stability, however, I would caution against the overly optimistic belief that “This time is different.”
Interesting perspective. I think this is exactly where macro analysis becomes important. A transformational technology can still produce disappointing equity returns if capital allocation, competition and margins move in a different direction than expected.
This Q is not directly related to AI, but what does the Yield Curve on Fed Bonds say about the economy. It was once the go to stat for predicting recessions and booms. Is it still relevant?
I think about the impact internet has had on our society and culture. Which I'd argue is negative. Great for business, great for instant communication and dissemination, but we've lost/are losing a generation plus of kids to screen addiction with related deficits in interpersonal relationship skills. Hardcore pornography, gambling addiction, surveillance technology, all these things and many more are problematic. Newspapers, books, radio are all media of the past and dancing with oblivion.
Change is inevitable. Not necessarily bad, but always hard on old-timers ( :/ when did I become one of those?). As someone who's watched internet technology crash into our society, I've long reckoned it will need maybe 50 years give or take, starting from the mid 90's, for the rules of the road to be worked out and systematized for best practices which reduce harm and maximize benefits of the internet. Buuuut... now there's A.I. supercharging the process.
My point is I think we would have been smart to adopt the internet's innovations more thoughtfully, where the need for guardrails could have been noted and flagged. A.I. is being pursued with reckless abandon. As someone who once enjoyed reading books a great deal, yet now find my time and attention captured more easily by this screen, the "need" for A.I. seems manuactured and very likely should have some thoughtful guardrails placed. We really don't have the capacity to understand the pitfalls until we've already fallen.
off topic
Lindsey Graham funeral shows that most corrupted and reviled politicians in America will get grand State funerals of a Hero. America is a bankrupt country.
America elected Donald not once but twice. That tells us a lot about America, and it is not good.
The oligarchs including Donald Trump control the Republican Party. Elon Musk, Peter Thiel, Jeff Bezos screw their workers and continue to block all legislation that will make housing, food, energy, health care, etc. more affordable. They are all cruel and hateful greedy bastards.
What I can't understand is why, as billionaires, they want more money. If you have a billion and spend at a rate of $5000 a day and assuming zero investment returns on your money, it takes around 540 years to run through it. Dunno about you but I could get along fine on $5000 a day.
They want the power that accruing more extreme wealth brings.
Harv Eker referenced the psychology of people who come into extreme wealth. For a few years, they typically go wild, buying expensive cars, clothes, planes, houses, etc.
But then they get bored, and look for a reboot. The psychologically healthy ones look for ways to instead help the country, the world. The twisted ones instead push for more wealth, more power.
They are broken people.
Oh the oligarchs spend far more than $5,000 a day. Just keeping your fancy yacht going will cost more than that
A close interior design friend of mine who works hard 6 days a week is thankful they do . . . it's truly "trickle down" economics.
The thread below has been hijacked, so regarding Dr. Krugman''s post: Financial markets have a mind of their own having nothing to do with us little people. Big money needs places to go and it's AI.
The stock market climbs a wall of worry and predictions of a crash seem greatly exaggerated. If it happens, that might have more to do with the stupid, metastacizing and endless war in the Gulf.
Can't disagree.
Just saw an apt line attributed to David Cochrane, a Canadian writer, in a comment following Jen Rubin's incisive Substack today: "America is not like this because Donald Trump is president. Donald Trump is president because America is like this."
https://www.contrariannews.org/p/trump-has-lost-more-than-the-iran/comments
I remember when I realized this, a while ago, and it was profoundly depressing. An election might help, but will not not solve what is wrong with America. Because it’s Americans themselves that are what’s wrong and that is not going to change in my lifetime.
See the movie, JF, "Miracle on the Hudson."
You'll be revived by the prompt, human response of so many first responders to Sully's passengers on that plane in the river.
Americans, I think, are OK. It's all those who did best on testing who learned only the most corrosive dehumanization to which schools have sunk.
Thank you. I remember that quite well, and the solidarity after 9/11. That was all pre-Covid. I think the pandemic broke America, because that served Trump well. Even by their very nature, pandemics are divisive and breed distrust; I remember David Brooks saying that right at the beginning of Covid, because it surprised me. Now it makes sense.
The people in "Miracle on the Hudson" and Trump enthusiasts could very well be the same, confused people. I do so hope that this confusion is temporary.
I think it also tells us about our schools! My parents could think logically, with basic training of how to apply their noodles. All of that type of education seems lost in recent decades. We must tune-up our children's education with common sense, civics and respect for one-another.
Why are you still a republican? I wouldn’t trust a single one.
This reminds me of my parents who first registered as Republicans in the early 1950s, before the party embraced the Southern Strategy of racism and religious extremism. By 1980 they were fed up and stopped voting for Republican presidential candidates, but remained registered as Republicans for the primaries in an explicit effort to hold the line against the crazies. It actually worked for a while in my state, but the nationalization of politics eventually won out.
I remember my parents, when moving to Florida in 1953 from Maryland being nonplussed at not being able to register to vote as Republicans..... How far we have come! (I'm not saying which direction.........)
Your language and arguments do not encourage belief in the "Republican Front".
That idea isn’t going to get you anywhere. It is too academic and doesn’t help anyone except maybe the senators.
The Republicans aren’t the problem. They are obviously broken, but it isn’t the core issue. Nobody is speaking up for the working man. Trump is the closest thing, but he isn’t going to lift a finger for them. He will stiff them on the bill like he always does. I’m all for the Republican Party becoming that progressive front, but I’ll believe it when I see it. The DNC is more economically right wing than the Republicans. They are fighting even harder to avoid it. The working people are mad, and until someone is doing something for them, they are staying mad, and we are going to be in this hellscape indefinitely.
And your battle cry is — "Start packing!" ?
halle burton – halli burton, I went strait into the language trap. Sorry, lol.
I was particularly struck by Vance's eulogy, where he recalled that he had approached Graham to vote for a particular project, thinking that Graham would lie, like any politician, but Graham told Vance he had already voted for it. Vance was surprised and told Graham he didn't know Graham was interested. "Oh I'm not," came the reply. "I did it because YOU wanted it." Vance was hugely impressed, he apparently thought this spoke to Graham's integrity, whereas it was merely the act of a weak man kissing the ass of a narcissist.
Pretty telling just how low the bar is.
Vance's lack of self-awareness always spills out in what he says.
Always.
I haven't looked at the morning paper yet; did Lindsey get a big sendoff? Weird guy. Weird country. It was weird in different ways years ago, but 21st century America is . . . words fail me.
Trump fell asleep sitting in the front row.
Surely he was just resting his eyes, or (gulp) thinking?
LOL, "thinking"... Yeah, no.
Perhaps the White House can spin it as prayer.
Forgot Graham's name during the eulogy. "Everyone should be like...this great person, sitting...lying, wherever..."
Too much early morning tweeting.
Zelenskyy flew in special to attend his funeral.
Graham was a supporter of Ukraine. And Zelensky I believe tries to do what's best for his country.
When Trump goes, tons of us should eat raisins so we can give a 21 bun salute.
LoL.
It was like a funeral for a Mafia don.
Face it fascism is in full force in the USA!
Makes one wonder what the obit scholars are preparing for McConnell.
McConnell is every bit the traitor that Trump is and is the cause of the politicisation of the US legal system.
His obit will read: "The senator continues to make progress in his recovery ..."
He is on Eternity Leave
Or a twist on Twain's comment: Rumors of my passing were prescient.
A real trifecta here - Smith-Prah-lemon. Funny.
Betting that there will be no mention of his outrageous Supreme Court shenanigans.
McConnell didn’t effusively grovel the way Graham did, so I expect some negativity will be allowed by the regime.
He was definitely sincere in his support of Ukraine. He took the train to Kyiv at least eight times to discuss matters and show his support. He pushed legislation to put more sanctions on Russia.
Graham was actually among the biggest lobbyist for US defense industrial complex - Trillion dollar hydra. Anywhere where US can market its merchandise, Lindsey was always supportive of that - He was a merchant of death.
And his alignment with Ukraine was just coincidental with his lobbying interests.
If US could sell its defense merchandise to Russia, Graham would've had 'sincere' support for Russia.
There was nothing 'sincere' about Graham. A sociopathic liar and a scoundrel of the highest order who deserves 7th circle of hell.
NSAlito….His “bads,” outweigh his “goods” by a long shot. He gave us the worst Supreme Court imaginable. He is a lying hypocrite. That is the bottom line.
I know that.
It's just that I only learned of his special support for Ukraine the day before he died, and I want to remind my brain that warmongering sniveling toadies aren't necessarily ·all· bad.
That's the GOP MO. Propaganda above all else.
I certainly wouldn't like the grand State funerals of a Hero that these corrupted and reviled politicians would get. It actually sickens me, but at least they'd be dead, good riddence.
Naw, you shouldn't talk like that.....speak only good of the dead: He's dead. Good.
Politicians who are veterans always have elaborate funerals. It doesn't hold much weight with everyone else. Just Graham's history in the Air Force increased his standing--or lying in state.
I recently saw the episode of Downton Abby where Lord Grantham bet on the wrong railroad & lost his wife’s fortune (whom he truly loved, but it helped that she had money when they married to prop up the estate's failing finances). The railroads were a revolution, but not all were successful, and it was hard to tell in real time how it would shake out. Lesson for today.
Important to note that Lord Grantham made the investment against the advice of his financial advisors. I believe they recommended more diversification.
I love that you use the term, "huevos". That's been one of my favorites for a long time. Is this commonly accepted forever? One of my first Spanish expressions was, "Dos huevos." I just commented about Andy Beshear and his missing "huevos"...
While I think AI is inevitable, I always hope for a “Dune” like system with human computers.
Our brains are now smaller than Cro-Magnon man&the Neanderthals, so maybe we should address that issue.
Because there’s money to be made&these electronic information systems are so pervasive&easy to use, societally AI is inevitable.
DATA Centers, I don’t know?? We really do need to take better care of our world.
Your last sentence is so important!
We're probably still a long way from weighing the benefits of technologies against the real costs of using them. It just never seems to be part of the equation to say, "this is really good stuff but we need to find ways of using it that minimize its impact."
Well that would interfere with maximizing the profiteering.
Welcome to capitalism! You are absolutely correct.
This isn't even 'really good stuff" - except for the entertainment value of watching lawyers get sanctioned for using AI in their legal briefs which hallucinate case authorities/quotations. (The med mal cases that have or may result are less amusing...)
Yeah, way "not funny!"
There is some really good stuff in niche areas but technology is rarely siloed limiting its use to things that have really favorable cost benefits. Actual environmental costs are not often considered.
The most important
We need to takr better care of our world and slowing waaaay doen on data center construction is part of that.
My in-laws are in their mid nineties.
They are so Republican I’ve had to bite my tongue more than once.
However, they protested a DATA Center in their community.
This reflects what so many pundits either ignore or don’t know: most of us have already listened to Republican-MAGA relatives during holidays and family reunions, and we’ve listened to similarly minded classmates and coworkers. Listening more isn’t going to solve the fundamental disagreement. We’ve already listened and concluded we want different outcomes and facts matter.
My Uber-MAGA cousin is protesting a data center in Florida!
Good LUCK!..............
rules for thee but not for me
NIMBY's sprout prolifically no matter, it seems, what political affiliation.
I’ve read enough science fiction to know this AI thing can be very dangerous. And then there’s too much information about our own personal thoughts out there for the wrong people to know about.
Agree
and flock cameras tracking and now capturing your electronic data.
deflock will tell you where they are and haveIbeenflocked will tell you if you are in their database
Apparently the critical thing is the amount of folding in the brain, not the size.
Men have bigger brains than women, and there is zero evidence that makes a difference.
While our domesticated brains are smaller, we have been able to use them more effectively by equipping them with more expressive language, education, and offloading cognitive load with technology like books (facilitated by printing) and computers. Our phenotypes have been really extended.
Back in the day, you had to be fairly sharp as well as attentive to multiply numbers written in the Roman numeral system. And it took a fairly long time to do so even if you could. These days positional notation, usually in base 10 make it a snap.
It is much easier to handle numbers with the Arabic script. Interestingly, my British handling of numbers for addition and subtraction is slightly different from teh way my son was taught in California.
Calculators are a huge boon to rapid calculations, and spreadsheets and computer programs vastly increase that facility. Yet when I taught university freshman math classes, I found that their reliance on calculators created what I called "number blindness". The students couldn't detect when they had made a mistake in a calculation. They had also lost the skills that I was taught in the 'slide rule era" to determine expected decimal places, create estimates of results using nearest 10s, and factor out large sets of numbers in divisions or equalities. All these techniques make handling some calculations easier, and ensure that the calculator answer is within th expected range.
So the calculator for contemporary students has meant that the cognitive effort that is now offloaded is also lost in the brain. This is the same as our ability to memorize prose and poetry has been largely offloaded into books.
This is something that makes the end of Bradbury's "Fahrenheit 451" so jarring. People preserving books by memorizing them word for word. As we know from the oral period before printing, this can not be done without error over generations, so that stories change over time. (It is like the Chinese telegraph game).
But our danger when putting texts into digital media, especially on cloud servers, is that it is prone to loss and deliberate manipulation. This is Orwell's warning in "1984" and the aims of the Party's Ministry of Truth. We can see the same desire in Trump's regime of trying to change history by removing historical exhibitions and removing or changing labels.
I know
@ Suki. Women's brains are on average smaller than male brains. Which shows that in some instances, size is not as important as it is in other areas.
Do we really know that absolute brain size is the deciding factor in the success of a hominim species? Neanderthals were exquisitely adapted as hunter gatherers and their brain size no doubt reflected that. Somewhere along the way, sapiens evolved brains with the capacity, among other abilities, to solve differential equations. Maybe (and I’m just speculating) that brain didn’t need to be bigger, just wired differently.
I forget the exact numbers - but it goes something like this: A single human brain can perform (a bajillion) calculations per second and only needs to consume about 1,000 calories of energy (a Big Mac, side of fries an maybe a couple beers.) An equivalent AI Data Center needs (umpteen thousand) gallons of water and about (some absurd number) of kilowatt hours to perform the equivalent number of gigaflops. My point: AI will be great for some things, but it's not going to recreate, all by itself, the Beatles Catalog. Let's leverage the strengths it offers but not worry too much about the World becoming a scene from Terminator.
Your math is wrong.
A modern single rack of Nvidia GPUs is far higher than the highest estimate for the human brain.
Needless to say, there are a lot of racks in a data center.
Perhaps 120 kW for that rack, but looking at kWhr in comparison to compute per sec is off by quite a bit.
What math are you referring to?
I’ve looked at thousands of brain images.
We’ve had computers in our home for 40years.
My husband has two engineering degrees&a masters in information systems.
The human brain has gotten smaller, our intellect has diminished&I don’t see any existing educational system doing anything to counter that bleak fact.
I’m aware of the capacity of the human brain.
I made a half humorous comment that you felt compelled to counter.
Don’t assume you’re winning arguments with me by attempting to school me.
I have a Commodore computer in the attic. Information systems evolve&exponentially are evolving.
I wish as in the “Dune” reference humans would make use of what we have.
Sorry. I was replying to Bill M when I said the math was wrong.
The latest version of substack notifies you when people under your comment reply to replies.
But for the record, Dune is science fiction.
Current GenAI may be able to reproduce Beatles catalogue, but it does not create anything on its own without a human prompting it. Not even a single word. GenAI systems have no thoughts, desires, plans nothing of that kind. Machines capable of that are still very much just sci-fi and I hope they will remain sci-fi in the foreseeable future (ideally forever).
@Martin. The systems imitate rather well. I recently asked Perplexity whether a certain garment made me look fat. In return I got a whole sermon about how I should not body-shame someone.
Again, my point is that if you did not ask Perplexity the question, it would’ve never said anything. The LLMs it uses have no intentions, desires, plans, opinions on their own. They can give expansive answers addressing aspects you did not explicitly ask about. However, you must ask (or at least say) something. Otherwise they stay quiet.
Generative AI can do planning. A recent example was Hugging Face incident:
https://www.bbc.com/news/articles/c3ek3gvdnj3o
During a test, an OpenAI model hacked into Hugging Face’s computers when searching for the answer. That shows planning. It also demonstrates, if not a desire to do well on the test, something close enough to a desire to do well on the test that the distinction may not be of practical importance.
My point is that the current AI must be given an explicit goal to do anything. It does absolutely nothing on its own. Once given a goal it can do planning, trial and error testing etc.
We understand this re AI.
I have had computers in our home for a least 40 years, I have a Commodore in the attic.
My husband thinks it should be somewhere safer, but my point is AI as it exists now is not where it will be in a very short time.
I’m not exactly in fear of AI, but I know the technology will evolve.
+I’ve seen all “The Terminator” movies.
Maybe we should do more with what we have.
That would be more cost effective than enriching tech.
Great episode of Downton and may very well be a cautionary tale for AI. We’ll see how all this shakes out. I would be curious about Warren Buffett’s initial take on this AI “revolution.” Thanks for your comment!
"It's the second mouse that gets the cheese..."
Apple has proven this forever.
I don’t know. What do you mean? Taking ideas that are out there and turning out a product that’s more functional is innovation, no? That’s not the same as eating off of a dead man’s table.
Apple used to be more innovative but lately they are not really first movers.
Since Steve Jobs passed, they have become more popularizers of ideas that are out there and proven than they are risk takers on the uncertain frontier. And they have been very successful at this.
Great episode of The Story of Money today about how the construction of the Erie Canal was financed with municipal bonds. Municipal bonds really are fascinating. Not all municipal bond issues succeeded, but I was surprised to learn how big the municipal bond market is, and how we will need them to finance all the infrastructure renewal that needs to happen. Great angle on how they relate to local democracy.
https://podcasts.apple.com/ca/podcast/the-story-of-money/id1376303362?i=1000778816657
Thanks for the suggestion. I added it to my list (with priority.)
The railroads are a great example of irrational exuberance.
I always feel the need to post this link to 'Agents of Chaos' in moments like this, because you are correct. AI, though, increases those risks because we're asking it to make decisions.
In this article, red team hackers set up a controled experiment in which AI agents with decision making authority are tricked into making bad decisions. One outcome in the logistics realm: they are able to shut down shipments of food to grocery stores.
We need to really absorb the damage we can do to our system by trusting AI too early.
https://arxiv.org/pdf/2602.20021
Downton Abbey’s storyline opens with the Titanic! Perfect metaphor for our time and yes that show was a lot of fun. I often compared myself to the ‘waitstaff’ at my last ft job, where I prepared communication materials for the company elites.
The railroads had a tremendous impact on the country and society but, as you point out, not all of them were successful. The same thing happened with the auto industry and probably with just about all new inventions. The problem from the point of view of the investor is -- which company will be successful and when (over what time period). Which AI companies will be successful and when?
Best wishes for a happy life.
Many a mother of a US girl with "new" money, worked hard to catch an impoverished Englishman with a title for said daughter. Edith Wharton started to write a novel about this phemomenon: The Buccaneers. (EDIT: this reply was meant to be posted elsewhere, @BKW.)
You read Krugman, too?
Huh.
We here all read Mr. Krugman's articles. Many of us also read the comments section. Some listen to the "musical coda".
Yeah, huh.
I just read an interesting piece on the circularity of investment between chip makers and AI producers that likened it to the model Enron used, which eventually led to the collapse of Enron. These guys are propping up the value of each others’ companies and the suggestion is that if any piece starts to unravel, it’s going to unravel quickly and it will be bad.
Indeed, pkidd! PK’s conversation with the AI expert at MIT, Paul Sodrowski, also mentioned this. And they’re using complex financial instruments like CDOs to finance their arrangements. Quite right! What could go wrong?
Say, I missed that talk and am having a hard time finding it. I tried searching the name in PK’s substack, but nothing comes up. Do you recall when it was? Is the name correct?
https://paulkrugman.substack.com/p/talking-with-paul-kedrosky
Ah, the ·other· PK.
Ah,I get it, now! 🙃
Just read this, think I may have understood some of it, especially the later parts on bubbles, etc. Will go back later and re-read if I can. But thanks very much for the link! And thanks, PK’s!
Thanks!
I love your handle. Whenever you reply to my comments, the notice says, “noone really replied to your comment…” There’s also a “Someone” out there. Other possibilities: someone actually, noone ever, noone ever really, just one a-hole, I, my dog, my next door neighbor (lucky you!), etc. 😂
Say, this is fun. Thanks for the laugh.
not to worry - they're too big to fail and the hoi polloi will bail them out
@Jim. As usual.
Regulation in Trump 2 is pretty much non-existent for every industry. As I read, all a company needs do is send an email to the White House.
Eh. It is not hard to tell.
1. Does the product they are selling produce tangible value?
Crypto consumes real world energy to calculate a random number on one end of the block chain with millions of computers, which could be calculated at the other end of the block chain by one computer in under a microsecond. I expect nuclear bombs are worse, but it is down there with military munitions for negative ROI.
AI has its moments for some tasks. So do spreadsheets and word processors. One would like to think that spending on AI would be commensurate with office suites, not national economies.
I now work part time as a parish administrator. For 23 years I worked for EDS (based in Plano TX) and HP after it bought EDS. I had a front row seat to the ENRON scandal as the project manager of the team charged with pulling documents for the SEC investigators.
I can tell you definitely: "the circularity of investment" generates a lot of smoke and mirrors that makes it really hard to identify what AI is really going to do... and also... who is going to do it.
Here is what I have been able to discern about AI so far...
1) AI is not going to take your job and is not going to make your job any easier. It will, however, change your job... a lot.
Take church administration as an example. Twenty years ago, people attended church in person, read the printed paper bulletin, donated with cash and paper cheques. Today, we want our personal preferences met. Those who prefer in-person still attend in-person, but they are only about half the population. Some (generally elderly) still prefer cash and paper cheques. So to reach the same people in 2026 as church did in 2006, we have to use video, Zoom, blogs, newsletters, electronic payment,... in addition to... everything that went before. Good that there is AI to transform content. But we in the parish office are running very hard in order to stay in one place.
2) The Tech behemoths of today will not dominate AI five years from now.
The AI implementations that solve real problems, for real people, will make real profit, and will eventually dominate. You will not know the names of these companies unless they solve a problem YOU have, as most of the problems will be very, very boring... such as better way to manage accounts payable, or scheduling the shipping of thousands of widgets from thousands of suppliers so that they all show up, at the same time, at a piece of land you've bought, in time for your contractor to assemble your house.
Moreover, along with contractor's hammer and spanners, AI will just be one of the tools. You will see the delivery vehicles and the guys with hammers. You will not see the AI that scheduled their arrival on site.
> The Tech behemoths of today will not dominate AI five years from now.
They don’t now. Anthropic and OpenAI are not in the Mag7. The behemoths have been giving them money, but so far they are not the same company.
"You will not know the names of these companies unless they solve a problem YOU have, as most of the problems will be very, very boring..."
----
Back during the tech boom of the 1990s, the advice to people who wanted to launch a tech startup was, "First ask the question: Is my product a vitamin or an aspirin?" This was an analogy to vending machines in the airport, which sold a packet of two OTC painkillers for $2, because when you need it, you need it. There is no market for vitamin pills in an airport.
Wow, thanks for sharing your very relevant experience. That's helpful.
You are correct to point out that the generation now graduating university has more mental health issues than their parents because they are more socially isolated. More zoom and social media won't help... enough. In-person church activities will help more. But they see in-person church as their grandma's activity, nothing to do with them.
If you take the statement above and paste it into NotebookLM, the AI will respond. Some of the web links and documents it finds will be useful to a church trying to develop strategies for church leaders over the age of 60 to reach people under the age of 30, such as your neighbor. That is an example of AI helping.
When I was an IT project manager I often had clients who thought they could solve their problem if they just threw more tech at it. I would tell them problems are personal: A problem stops SOMEONE from doing SOMETHING that they want to do. If you cannot identify who to the point you call them on the phone, and what in enough detail you can put a dollar value on being able to do it, more tech is not going to help.
This is the situation of AI today: Too many executives expect AI to solve their problems without them having to do any hard thinking.
Maybe she doesn't want to "pair up"?
Video meetings mean participation is possible when you are disabled, and travel is very difficult.
Are you talking about Ed Zitron's piece from yesterday?
https://www.wheresyoured.at/the-more-you-buy-the-more-you-lose/
Was that Dean Baker?
https://cepr.net/publications/ai-bubble-monitor/
or Ed Zitron?
https://www.wheresyoured.at/the-subprime-data-center-crisis/
Another vote for Ed Zitron. Been following him since I stumbled onto his work the first of the year. It could stand some organizational and copy editing (and fewer f-bombs), but it's thoroughly researched and a blunt, clear-eyed critique of the AI media-and-finance frenzy. If he's right, it's not going to be pretty, and it could resemble the dot-com and housing bubbles combined, overlaid by hundreds of billions of Enron-grade financial shenanigans.
It wasn't but thanks for sharing those links - - here's a link: it's on "Notes from the Circus" (Mike Brock) https://substack.com/home/post/p-208866925
Annoying art historian here: Is the cat's face actually "buried near its front paws"?
My thought exactly.
Exactly. Back in the day there was 'Nicolas Cage is a ripe pomegranate' trope about classifiers, namely that for any input they were obliged to produce some sort of output, and usually drawn from a smallish set options at that. Hence the caution that a classifier built to assess the ripeness of a fruit simply does not have the option to output 'Nicolas Cage' even when the input is, in fact, an image of Nicolas Cage.
The trope still holds, no matter how much more complicated and powerful the machinery is today: By its very nature, the product currently being hyped as 'AI' can only interpolate; that same nature guarantees it never extrapolate outside the range of the data it was trained on.
Technically yes.
Amazing; AI has already learned "politician speech" fuzzy language.
You are right: back paws are closer than front paws.
Also, the bed cover is not a patchwork quilt - the pattern is printed/on the cover. No big deal when considering the impressive discussion of the cat's photo, but a similar mistake in, say, a legal document (where AI is expected to earn its keep) could be a deal breaker.
It is not. Still impressive, but they'll never get rid of the lying.
Maybe someone should ask Claude and ChatGPT and Grok "What does the future of AI look like?"
Hear, hear!!
And if they all reply: "I'm sorry, Dave.."
That's funny.
I use three AI apps for comparison and translations. I give them the same text in English and ask for a German translation. Then I ask them to translate the text to English. Not one of them comes back to the original, and all three give different results.
.
Once tried to get a company to translate an operations manual into Chinese (was working there) and then get a different company to translate back into English. I did not get permission - I suspect we would have had a hard time understanding anything from the 2nd translation
I once owned an Italian motorcycle. I think the owner's manual went through at least two other languages before making it from Italian to English.
Please not put the before item here.
There are stories -- most of them apocryphal, no doubt[1] -- about early translation programs going from English to another language and back again so that, for example 'Out of sight, out of mind' came back as 'Invisible idiot'.
[1] https://www.snopes.com/fact-check/quality-is-the-first-occupation
That was a game on a CBC radio show, Lost In Translation (IIRC), try and identify the song from round robin translation of lyrics.
E.g.:
You’re wondering who I am, aren't you? (Secrets, secrets, I’ve got secrets)
Am I a machine, or a mannequin? (Secrets, secrets, I’ve got secrets)
Made of Japanese-made parts (Secrets, secrets, I’ve got secrets)
I am "Modern Man"
oʇoqoɹ ˙ɹW - xʎʇS
😂
Seems like you forgot to have a control group.
I expect you would have the same result if you asked humans to do the translation, at least assuming the text is not trivial.
Remember, "English as she is spoke"? Human translators are not any better.
edit: hear = heard
Hmm... no, I never hear that. Was it taught in your school?
"English as She is Spoke" was first published in 1855 by a Brazilian who took a Portuguese-French phrase book and a French-English dictionary and created a Portuguese to English phrase book. English missionaries found it in use as a text book in schools where they went to teach. Today, you can get it on Amazon, but in the "humor" category.
Thanks for the explanation. No surprise that it's not really known in Germany. 😊
Translators are traitors
Forn languages are jus' suspishus.
(You can be like Hegseth, and only speak ·American·.)
Today in the NYT there is a description of efforts to make an AI do the work of an administrative employee. Didn't go well, but I suspect a new human employee might have problems too until they get the hang of a job.
Meanwhile, dealers in rare and antique books are reporting that our tech lords are buying up old books to be ripped apart and fed into the scanners for AI. These people respect nothing, least of all you and me and what we might find a desirable future. Mars anyone?
For the tech bros, presumably? I think we should facilitate their departure asap.
Anyone else read the interview with Zuckerberg in the NY Times? The vitriol directed at Zuckerberg in the reader comments was something else. And it was comment after comment after comment. I guess the tech bros have fallen down a few pegs. Cue the tiny violin music. https://www.nytimes.com/2026/07/28/technology/mark-zuckerberg-meta-ai.html
The dealers really aren't in a position to complain. Taking apart old books and selling the individual pages as "art" has been going on for at least 50 years. I still have a few pieces that my mother bought in Europe in the '70s. Most have certificates of authenticity.
Vernor Vinge's sci-fi novel, "Rainbows End" has a subplot of the protagonist upset that his university library was being fed into a "chipper" and the fragments digitized and reconstructed into digital books. (A bit like Venter's shotgun genomic reconstruction for texts.)
I too am appalled at the idea that old books are destroyed to make it easier to and cheaper to digitize them. OTOH, this has been done before for a different purpose. Back in the 1990s at a San Francisco Book Fair, one could buy individual illuminated pages extracted from old religious texts. I don't even like buying used books with writing in the margins and highlights.
that's truly awful
Nathan Myhrvold, ex-technology chief at Microsoft, was well-known for buying very large numbers of books each year. Far more than he could read, I am sure, but we cannot know that he didn't read some.
I still prefer reading physical books, but I probably read far more online, like this blog.
The massive expenditure on "AI" infrastructure has got to be the greatest misallocation of capital in world history. And that's to say nothing about the opportunity cost. At these prices, we could have had a full upgrade of our power grid and electrical generating facilities, for example.
AT&T used to be the phone company. They were broken up and the resulting company left with the name went on to build several amazing businesses that they always seemed to start a little too early and were never able to be successful. The assets would get sold off and someone else would make a mint off of what they built.
I wonder if the frontier AI companies or maybe those dumping billions to not catch up to them will serve the same purpose. Will Oracle or Facebook or SpaceX be left with all of this capacity but models that don't measure up to the market? Will other companies be able to take their capital and build the AI equivalent of Google that turns the equivalent of AOL or Yahoo into a zombie while the upstart thrives?
That Trump wants to ban Chinese open-source AI models is indicative of the threat they offer to big AI Labs like OpenAI and Anthropic. I have just read that the US is going to ban Chinese humanoid robots that seem to be advanced and much less expensive than US ones.
Love the cat! :)
I believe that's Jack, research assistant. Tuxedos are some of the sweetest cats!
Unlike people, cats care not at all about fur color.
I wonder where the money will come from for the millions of people unemployed because of AI, to live their best lives sipping margaritas by the pool, after a hard day's shopping? Do you think the tech bros are saving up for this? Who will pay for your house? Your car? Your child's education? Elon will feudalize you; you will be scrounging in the dirt for a slug to eat.
They'll be hired back before long, because the "AI" will fail to deliver on the hype.
they will be doged - fired then rehired
If we do the wealth distribution right then we can all retire at age 55 and spend our time doing more fun stuff. You know share the work and share the wealth.
LOL
The episode with OpenAI seems likely to have a greater effect on the market. It is absolutely scary.
After reading all of Burry’s analysis, and then hearing Sam Altman say, “We’re going to let our AI figure out how to pay for itself,” I’m feeling like the future ain’t too bright.
GIGO
As a Korean, I feel deeply saddened and concerned about the vulnerability of the KOSPI and KOSDAQ.
Finance, by its very nature, inevitably involves risk-taking in the pursuit of returns. However, Korea’s capital markets are relatively small compared with major global financial markets, and the scale and speed of the recent market turbulence have been truly alarming, given their limited capacity to absorb such abrupt volatility. I remain optimistic about advances in AI technology. From the perspective of financial stability, however, I would caution against the overly optimistic belief that “This time is different.”
Interesting perspective. I think this is exactly where macro analysis becomes important. A transformational technology can still produce disappointing equity returns if capital allocation, competition and margins move in a different direction than expected.
That cat is like get Claude away from here.
This Q is not directly related to AI, but what does the Yield Curve on Fed Bonds say about the economy. It was once the go to stat for predicting recessions and booms. Is it still relevant?
I think about the impact internet has had on our society and culture. Which I'd argue is negative. Great for business, great for instant communication and dissemination, but we've lost/are losing a generation plus of kids to screen addiction with related deficits in interpersonal relationship skills. Hardcore pornography, gambling addiction, surveillance technology, all these things and many more are problematic. Newspapers, books, radio are all media of the past and dancing with oblivion.
Change is inevitable. Not necessarily bad, but always hard on old-timers ( :/ when did I become one of those?). As someone who's watched internet technology crash into our society, I've long reckoned it will need maybe 50 years give or take, starting from the mid 90's, for the rules of the road to be worked out and systematized for best practices which reduce harm and maximize benefits of the internet. Buuuut... now there's A.I. supercharging the process.
My point is I think we would have been smart to adopt the internet's innovations more thoughtfully, where the need for guardrails could have been noted and flagged. A.I. is being pursued with reckless abandon. As someone who once enjoyed reading books a great deal, yet now find my time and attention captured more easily by this screen, the "need" for A.I. seems manuactured and very likely should have some thoughtful guardrails placed. We really don't have the capacity to understand the pitfalls until we've already fallen.