Skip to main content

Posts

Showing posts from August, 2024

The LLM Bubble is bursting

Some observers, including analysts for Goldman Sachs, say that the growth of AI since the height of the pandemic is now proving to be a bubble, like dotcom stocks in the late Clinton years or mortgage based derivatives in 2006-07.  What is a bubble? By standard definition it is a growth in value of an asset that occurs after the speculative demand has surpassed the operational demand for same. If I buy X as a speculator at a price higher than the operational value of X can support I may be willing to make an operationally foolish deal in the expectation (on the speculation) that there will be a "greater fool" to whom I can then sell it.   In such a situation, sooner or later the biggest fools ARE the ones who are holding X. What happens to its value then you can fill in for yourself.  Clearly there is money still entering the field.  The Peoples Republic of China is incubating a group of companies that it proudly calls the "little AI dragons," with seed money of mor

"All nulls are false!" Huh? Deborah Mayo

As you probably remember from my frequent recent mentions, I've been working my way through a book on probability, statistics, and the philosophy of science, STATISTICAL INFERENCE AS SEVERE TESTING, by Deborah Mayo.   This will be my last quote from that book. From p. 367, "It is supposed in many fields of social and biological science that nearly everything is related to everything: 'all nulls are false.' Meehl dubbed this the crud factor." What follows is a description of a psychology survey done by the University of Minnesota in 1966 in which all of the following showed some degree of (often low though statistically significant) correlation -- attitudes toward school, leisure activities, educational plans, parents' occupations, siblings, birth order, family attitudes. religious preferences, MCAT scores, etc. Every null hypothesis is false. Mayo tells us that Meehl here is also eager to explain that this is not what statisticians call a "Type 1 error&qu

The Hillbilly and the selfish genes

In 2020, JD Vance appeared on a podcast hosted by Eric Weinstein. Here Vance uttered the "yes" heard round the political world. The two were discussing grandparents. (And yes, we can all agree that the relationship between grandparents and their offspring's offspring can be a wonderful one.) But Weinstein wasn't going to let the matter rest there. He said that raising grandchildren is "the whole purpose of the postmenopausal female." Vance's response to THAT? A single word. "Yes." Indeed (for context), the whole Weinstein sentence, is "That [child rearing assistance] is the whole purpose of the postmenopausal female in theory." One thing I find fascinating about that wording is that Vance actually interrupted the sentence, to say "yes," BEFORE the host got out the phrase "in theory". Thus he suggests that for him anyway what he is agreeing with is not just a theory but a settled fact. Hmmmm. So for those p

From a screed re: the internet of today

It isn't just your grouchy imagination. Everything WAS better in the 1990s, before the "enshittification" of the internet. Or so says a delightfully written screed on the subject, "Disenshittify or Die," by Cory Doctorow. I'll just quote one little detail as a sample of the vigor of the writing. "The last time Congress managed to pass a federal  consumer privacy law was in 1988: The Video Privacy Protection Act. That’s a law that bans video-store clerks from telling newspapers what VHS cassettes you take home. In other words, it regulates three things that have effectively ceased to exist." But the golden age of the internet came AFTER that.  We lost our ability to regulate enshittification before the golden age even happened. That's the hypothesis.  Want to read more? I'll just link: https://pluralistic.net/2024/08/17/hack-the-planet/#how-about-a-nice-game-of-chess   BTW, the line in that URL for this screed, "how about a nice game of

New discussions of emergence

I've discussed the concept of emergence here repeatedly.  It is central to my still-developing personal philosophy, as it centers the mind-body problem and indeed the place of the human species within the broader world.  Not minor points.  Emergence is the idea that reality is layered, and that one layer arises as the simplifying embodiment and consequent of the complexities of the level that lies beneath it. One notable example of this is the emergence of consciousness from the complexities of life.  As it happens, there has been much discussion of emergence in this sense among contemporary philosophers. I will say a little bit today about three philosophical names to contend with: Jaegwon Kim, Sydney Shoemaker, Warren Shrader.  Jaegwon Kim (1934 - 2019) argued against emergence in the sense I've discussed it. He called it "non-reductive physicalism" and his bottom line was that it cannot be maintained. Physicalism is the future of philosophy, but it must accept a re

A long-abandoned favorite: Scott Adams' loop theory of the news

I haven't done this since 2016. I used to work it in fairly often: once a year or more. Scott Adams, of Dilbert comics fame, said that we're in a loop, in which a small number of news stories is endlessly repeated. As if our universe is The Matrix, and its coding economizes by repetition. Making this argument, he listed nine headline you'll find in tomorrow's newspaper.  1. EXTREME WEATHER BATTERS SOMEPLACE 2. IDIOTS KILL INNOCENT PEOPLE 3. POLITICIAN DOES SOMETHING ILLEGAL 4. PRIMATE ATTEMPTS INAPPROPRIATE SEX 5. EXPERTS WARN OF FINANCIAL CALAMITY 6. BIG COMPANY BUYS ANOTHER BIG COMPANY 7. FAMOUS PERSON DOES SOMETHING INTERESTING 8. A SCIENTIFIC DISCOVERY MIGHT BE USEFUL IN TEN YEARS 9. GOVERNMENT FAILS TO ACHIEVE A GOAL I'll match these up against real news, trying for some geographical diversity as we go along. 1. A hurricane just hit Bermuda.    2. Idiots who kill?  A lways too easy.   3. Politician does something illegal? Today's example (just for the sake

Sci Fi and burlesque parody

  Worth recording.  The Empire Strips Back: A Burlesque Parody - Albany - Tickets | Fever (feverup.com) Can't make the trip for the sake of the show but the advertising is amusing.  And boy am I getting lazy as a blogger. Tomorrow will be the result of some laziness too, according to current plans, though the fourth item this week will involve something a little more substantive.

Power as a statistical term

Just a quick vocabulary note by way of blog entry today.  In statistics, the word "power" is used as a synonym with "sensitivity". The test of a hypothesis has POWER to the extent that it is likely to detect an effect if it exists.  Typically in scientific experiments there will be a null hypothesis and an alternative hypothesis of this form: the null is that there is no relationship between two variables, the alternative is that there is one.  If a powerful test fails to give the tested-for result then, by definition, it supports the null hypothesis. As you may already have determined, I continue to make my way slowly through the book by Deborah Mayo that I have mentioned here a couple of times. Boning up on the vocabulary of the field as I go.  The above diagram should help explain the significance of power so understood. The further away the modes of the two bell curves; the smaller the possibility of an erroneous decision.

The revolution in Bangladesh

The recent events in Bangladesh have passed almost unnoticed by the 'mainstream' press in the United States.  The gist of it: a successful revolution seems to have arisen out of a university-centered "quota reform movement".  Here is a primer.  The Presidency in Bengladesh is a quasi-monarchical institution on the British model, most ceremonial. Muhammad Shahabuddin was made President by parliament about a year and a half before the revolutionary events discussed below, so he now represents continuity.   The law in Bengladesh 30 percent of government jobs to the families of the "freedom fighters," those who were active in overthrowing the "East Pakistan" government and creating modern Bangladesh in 1971. Revolution The quotas caused students outside of the charmed circle to believe that their prospects were limited. A decision by the Supreme Court of the country upholding the 30 percent rule set off the protests to which the government reacted in d

Is there a "just noticeable difference"?

Working within a branch of experimental psychology called psychophysics, scholars argue about "just noticeable differences," thresholds of perception, and related postulations. Consider the volume of sound as an issue. It seems intuitively there should be sounds that are different from one another in an objective sense (mechanically measurable) in that one is louder than the other, but that are not perceived as different by human beings.  After all, we did not evolve as mechanisms for the precise measurement of sound,  We evolved, to be simple about it, to survive and reproduce. That does not require ideally good distinctions in these matters.  So there should be some just noticeable difference between sound A and sound B as to volume, such that if I hear any pair closer in volume to each other than those two, I will perceive them as identical. Right? Likewise with weights? Put one object in my left and put another in my right and ask me to tell you which is heavier.  At some

John Dewey on education

Five points on Dewey as an educational theorist.    1. Dewey tells us that the key intellectual purpose of education is the development of reflective or critical thinking as a skill and a habit. Even without decoding the adjectives, we can see that the emphasis here is going to be on THINKING, an activity, not a passive body of knowledge that has to be passed along like a family heirloom. 2. But let us get into the adjectives. Reflective thinking is the habit of actively, persistently and carefully considering any belief or supposed form of knowledge in the light of the grounds that support it, those that suggests doubt, and the further conclusions to which it would lead it adopted. 3. Memorizing data so that one can recite it back again on command is not reflective thinking. It has its uses, but it should be some distance from the center of our thinking about thinking, that is, from a system of education. 4. Dewey was actually a critic of the "child-centered approach" to ed

Myron Kriegman

 Myron Kriegman !  I had a moment recently in which I had been blocked looking for the name for a certain character in the John Updike novel ROGER'S VERSION (1986).  Now I am determined not only to remember the name but to make it the subject of a blog post. So ... here it is!   I no longer have the book around: I may have donated it to a library at some point in recent years. The novel turns largely on a three-way argument over God and the worshipper: whether there is a God and whether that question can be answered by reason or requires faith. Kriegman occupies the "no to God -- and yes to reason" position in the triangle. The titular Roger (last name Lambert, although we are supposed to connect this to Roger Chillingworth, from Hawthorne) occupies the "yes to God -- and no to reason" PoV. Roger sees himself as a follower of Karl Barth in believing in a hidden God, who cannot be reached by human ratiocinations.  The third character in the triangle is Dale, a na

Kenya's bond rating

I write today in order to give the government of Kenya some credit for creativity in the face of adversity.  This week a key bond rating agency, Fitch, downgraded the sovereign bonds of Kenya, from B to B-. As influential rating agencies go, Fitch is on a par with Moody's on the one hand and Standard & Poor's on the other. As it happens, Moody.s downgraded Kenya earlier this year. The S&P may announce a grading change on August 23d. The change from B to B- is not a plunge into "junk bond" territory.  What are, sometimes unfairly, called junk bonds are more generously called "speculative bonds" in contrast to the widows-and-orphans are safe with these ("investment grade") bonds. The line between investment and speculation is, on the Moody's system, the line between triple B and double B. Accordingly, Kenya as a single B country, was already in junk bond terrain. The addition of a minus sign is an incremental indication of greater odds of

A thought on the presidential election of 2024

  During the now-infamous debate between the former President and the incumbent President, the debate that forced the incumbent, Joseph Biden, out of the campaign altogether, there was a moment that looks a lot different now than it did at the time. There was a moment when Donald Trump was going on about the wonders that would be worked for America by his, Trump's, great relationship with Putin. He said, for example, that by a simple phone call he, Trump, would get the Wall Street Journal reporter, then being held in prison by the Russians, released. In fact, Trump even spoke some kind words about that reporter, Evan Gershkovich.  I don't remember the words -- and I'm too lazy to look them up-- but they were an exception to the general rule of calling reporters "enemies of the people".  As is his wont. Except, apparently, when praise is convenient. Anyway ... it would only take one phone call....  At the time, Biden simply kept his head down and stayed silent. Th

Que sera, sera

The full lyrics here.  Comment below. When I was just a little girl I asked my mother, "What will I be? Will I be pretty? Will I be rich?" Here's what she said to me "Que será, será Whatever will be, will be The future's not ours to see Que será, será What will be, will be" When I grew up and fell in love I asked my sweetheart, "What lies ahead? Will we have rainbows Day after day?" Here's what my sweetheart said "Que será, será Whatever will be, will be The future's not ours to see Que será, será What will be, will be" Now I have children of my own They ask their mother, "What will I be? Will I be handsome? Will I be rich?" I tell them tenderly "Que será, será Whatever will be, will be The future's not ours to see Que será, será What will be, will be" ---------------------------------------------------------------------------------------- This is not a bad account of one aspect of ancient Stoicism.  What y

A thought about Gandhi

I've been reading recently about Mohandas Gandhi -- chiefly because I accidentally encountered a fascinating article by a fellow named Eljiro Hazama, published last year in the journal MODERN INTELLECTUAL HISTORY, a journal published out of Cambridge University. Hazama says the usual account of Gandhi's philosophy (with its two key components, satyagraha and ahimsa ) is a myth. Satyagraha means something like "force of truth" in Sanskrit. Ahimsa means "without killing," and usually refers to dietary and ritualistic matters. The "myth" that Hazama is complaining about consists of the view that these are old Hindu religious principles that Gandhi learned in the course of a religious education as a child, and that his life consists simply of a series of applications. Hazama argues that Gandhi nowhere uses "ahimsa" in the political sense of "non-violence" as an element in a political philosophy until he is in his mid 40s. He used

Deborah Mayo and particle physics

Readers may recall that I've been writing of late about Deborah Mayo, a professor at Virginia Tech.  To restate, Mayo has written Statistical Inference as Severe Testing ; a book discussing statistics, induction, and the philosophy of science. Much of the book is obscure to me, though I skim through what I have to in order to get to stuff where I can gain something. To move forward nonetheless: near the middle of the book she has a discussion of the Higgs boson. Allow me to talk about that today.  Back in 1964 Peter Higgs predicted the eventual discovery of a certain particle that would be responsible for other particles having mass. This became known as the "Higgs boson" or, more grandly as the "God particle"! Over the following decades the Higgs boson came to be regarded as an important piece of the Standard Model of particle physics, and the fact that no one had experimentally confirmed its existence became something of a scandal.  Then came 2012, and CERN,

The future of "cookies"

 Cookies, in their simplest form, are an algorithmic device for helping websites record information about their users. They are supposed to enhance the user experience by allowing for personalized content. They also help advertisers (and thereby help website owners) by enabling targeted ads.  Naturally, then, they have come under scrutiny on privacy grounds.  In the European Union, there is a General Protection Data Requirement that requires websites to get explicit consent from users before collecting any data by means of cookies.   Yet that can't be the end of the dispute.  At least three factors seem to be driving us toward a cookie-less future: Consent fatigue:   With Users quickly tire of the pop-ups requesting their consent to the cookies. It comes to seem like an artillery barrage. Technological limitations:  Cookies don't give the advertisers as much value as they'd like. After all, they are device specific, and much of the web browsing public moves their browsing b