analog31

Fortunately proofs were still part of the curriculum when I was in school. Typically, geometry focused on proofs. When my kids took geometry, proofs took a back seat to problems. This also got us to the point where the college math curriculum has to include a class on how to do proofs.

It's certainly possible to automate experiments. In fact my job is developing equipment for this purpose, and all of my own experiments are controlled by code. Most measurement instruments have run on code for decades. You don't manually turn the knob on a Wheatstone bridge to measure a voltage any more. A yellow handheld device with a LCD measures it with greater accuracy, or a probe that plugs into a computer. The code living in an experiment like the Large Hadron Collider is extreme.

The distinction between theory and experiment is somewhat arbitrary. Historically, scientists were better at one than the other, with a few rare people being known for both. To make an analogy, there's also no immovable boundary between hardware and software, since most hardware runs on code, but these tend to be different jobs and even departments.

What tends to happen in science is that when a subject is virtualized, e.g., when it can be modeled by a satisfactory theory, scientists move on to new frontiers. Many of the most interesting areas of work are where both theory and experiment can challenge one another.

I've had managers and even engineers ask me in all sincerity why we still need to make measurements when we can just model the data.

Oddly enough, one of my upper level math textbooks (maybe real analysis, don't remember) said the same thing about the definition of a set, except that "later" meant graduate school.

When all you have is a hammer, everything looks like a nail.

I refer to what I think the author is asking for, as "push button reproducibility," i.e., the idea that the results will reproduce themselves at the push of a button, anywhere, at any time in the future. I have a couple of misgivings about this.

First, the whole idea of "open" research predates computer technology. Forcing science to keep up with the latest ideas in software distribution is too much of a burden, when science is already too risky and slow. I had the odd privilege of learning the scientific method from my mom, before there was widespread access to computers. Her version was that a study should be reproducible by a reasonably skilled person. This is a greatly relaxed standard, but is realistic for a discipline that spans decades if not centuries.

I supplied all of the data and code for my thesis research (and a sufficient number of mechanical and electrical drawings). But nobody has Turbo Pascal today, and some of the commercial instruments such as specialized lasers were already obsolete by the time I finished. Also, the experiment was dangerous, and might not pass safety review today. It required about $500k of equipment and a dedicated lab.

Today I have the luxury of saying that if my code fails upon loading a new version of a dependency, the person who discovers that failure is probably skilled enough to fix it, and my work rarely hinges on the idiosyncracies of dependency versions. If it goes into a product, they'll totally rewrite it anyway.

Second, science is still at its core an experimental discipline. Even in physics, there are more experimentalists than theoreticians. Reproducibility means roll up your shirt sleeves and head for the lab. To this day, some processes have not been mechanized, and you still need to spend years developing "lab hands" which not all people succeed at.

I think there's a clue in the fact that the social and medical sciences seem to be the most deeply embroiled in the reproducibility crisis. It's because the quality of results depends on the the quality of measurements, and it's just harder when dealing with living subjects or one-of-a-kind specimens (such as the earth's climate). In fact, not much more than a century ago, it was believed that studying those things was beyond the reach of scientific methodology.

Third, we're not going to stop doing science in areas where it's hard, particularly in medicine, but we're also not going to staff up in areas such as software development, to make science work better. People are suffering from disease right now so there's always an urgency to finding cures, plus an obvious profit motive.

Disclosure: Experimental physicist, developing better measurement equipment.

Ironically, Max Planck's funeral was in 1947.

Disclosure: Old scientist.

The fear may be outdated.

I'm 62, and grew up near Detroit. I remember cars developing major rust holes after just a few years. The rusting also affected the inside of the engine compartment, and it did make repairs more difficult. There was an industry of aftermarket rustproofing shops, that were a scam. Every teenager with a car (which was still rare) learned how to patch rust holes with Bondo.

My first car, the frame rusted out and had to be welded back together.

Toyota and Honda put a stop to that. They did a number of things: Better steel alloys, better coatings, painting the bodywork before it could start rusting, eliminating places in the body where water could accumulate, etc.

Today, you rarely see totally rusted out cars on the road.

Oddly enough my dad worked for a chemical plant that mined salt that way.

An interesting aside is that the process for making these salt domes is also used for mining salt. Drill a hole, pump water down, and up comes brine. The brine can be used as a chemical feedstock, e.g., for production of things like chlorine, sodium hydroxide (lye), and sodium carbonate for glass.

The odd thing is that we do train people to do real physics, somehow despite my cynical take. I see the education process as following heuristics that are believed to have some relationship to developing real world abilities, even if we don't know why. We all worked these problems, and now we're somehow able to do physics. In the end we don't know what turns people into physicists. Or musicians, artists, etc.

An analogy is making students write 5-paragraph essays. We don't really believe that the knowledge of how to write a 5-paragraph essay is a real-world ability, and the LLM's can write them all day long. But for some reason we believed that 5-paragraph essays were a good pedagogical tool.

But whether the same heuristics apply to teaching an LLM is anybody's guess. If they learn differently than we do, then feeding them on our learning tools isn't necessarily going to help them.

I'm also a trained physicist, spent some time in industry, and taught a college freshman math class for a semester a couple decades ago.

There's an interpretation that makes the reference answer correct. I didn't realize this until halfway through teaching the math class: Based on the lectures, textbook, and homework problems, you match the "form" of the problem with similar problems in the textbook, and apply the same algorithms to solve it. The students have some vague idea of this, referring to it as finding the "trick," but it's never explicitly explained to them.

I was glad for the interesting puzzles and good grades. But physics really came alive for me in the lab, where mother nature decides the conditions of the problem being solved.

And the seat in front of you shattering your knees.

Modern materials and production methods eliminated the need for ornamentation to prove a craftsman's skill and attention to detail.

I'm thinking of something like a violin. You can't pick up a violin and judge the perfection of the fitting between the wood surfaces that are glued together. If that work isn't perfect, the instrument will eventually fly apart. But you can look at the "scroll" and "bees sting" which are unforgiving of tiny errors. These kinds of features persist to this day due to traditionalism, but have disappeared from most industrial products.

Today, I don't expect an ornamented house to be of any better quality than a plain house.

Indeed, the beige boxes are neither modernist nor classicist.

My family lives in a MCM house built around 1960. We bought it in a "hot" market where you couldn't be particular about anything. We chose a house based on location and the interior.

This is the "two truths and a lie" methodology.

I have music running in my head most of the time, at least in the background unless I’m concentrating deeply on something. I’m also a musician and the best way I can describe it is that the music I’m hearing in my head somehow comes out of my instrument.

I know, it's a different world. I'm 62. As a kid, my friends and I chased one another around the neighborhood with toy guns for hours, raising a racket. My kids are 24 and 26 now, but it would have been unthinkable to let them have toy guns.

>>> What if Waymo thought the guns were real, called the cops, and the cops screwed up and shot one of the kids? These sorts of escalations aren't imaginary; they happen, and they end in tragedy.

I take this seriously, but I think this a problem with the cops. Even regular people hesitate to report problems if they think the police might overreact.

I wonder about bakeries too. When traveling in France and elsewhere in Europe, the uniformity of bakery bread has suggested to me that a lot of it is made from a mix.

That's a fair point, but I think real learning is still worthwhile, and employers have always had to vet candidates beyond just the credential by itself.

The cheaters are supporting the students who choose to learn. In the past, this was the role of the rich kids who only needed a "gentleman's C"

The above post mentions big band on double bass, which is precisely my main bag. This is just to clarify. The sheet music consists of portions that are fully notated (i.e., dots on the staff) and lengthy sections with nothing but chord symbols. I'm expected to improvise those parts, and to take liberties with the notated parts as well.

So it is indeed a mixture of reading and improvisation. But this is aside from the choice of notation, which is just a practical matter of finding people who can read it fluently enough to perform from it.

Indeed, the musical traditions of western Europe and Africa, forged together in the Americas, dominate world culture. Turn on the radio anywhere in the world and you will hear the influences of jazz, rock, funk, samba, hip-hop, etc.

Indeed, in the typical progression, a music student is a fluent reader by the time they take a theory course. Reading was probably part of their entrance audition.

Maybe their needs to be a “music theory for engineers” like we have “physics for poets.”

Depends on your stats class. The one I took in college started from first principles, and was focused on proofs.

I’m a jazz player but not a theory expert. A lot of what jazz is about is rhythmic, and the elevation of improvisation to a very high level.

I know a lot of people with music degrees. This is a widely acknowledged problem. It doesn't have a simple answer, especially as academic music programs are shrinking. You can't teach everything except at a survey level. You have to find faculty who can teach new topics at a college level. And students who are prepared to dive in. Music students tend to already be damn good musicians by the time they start college.

If programming is an analogy, it's said that once you've learned your first language, you can learn any other language in a jiffy. Also, a lot of higher level learning is "learning how to learn." Maybe it's better to have a standardized "first language" that students can learn in depth, and then let them move into other areas at their own discretion.

I don't have a college degree, but for me, classical music lessons were a great preparation for learning jazz, and I've picked up pop, rock, and folk music on my own. In fact, on the instrument that I play, the double bass, gaining proficiency and experience in jazz without that background is a brutal learning curve.

No. The New Yorker never gets to the point.

— Long time subscriber.

As an aside, learn to properly care for cables. There are several YouTube videos on this topic, typically from musicians. Most importantly for small molded cables, avoid sharp bends and stress on the wire near the ends. Just a loose coil is much better. And the “over under” method for long cables avoids getting them twisted up.

Even in the past a lot of “aristocrats” had to earn a living. Among scholars, many worked as teachers or had sinecures that paid the bills, or held clergy posts with minimal responsibilities.

Aristocrats were a dime a dozen.

Falling below subsistence means people starve and cease to exist until supply and demand equalize or “something else” happens. That’s what Smith was getting at. Thus subsistence is a price floor for human labor.

Something else could mean developing a new concept of ownership or a new form of redistribution.