Novels as models, and ChatGPT

by  Tyler Cowen

I read your piece on novels as models many years ago, and I’ve been reflecting on it with the advent of LLMs. I wrote a piece (substack) making the case that the data required for AGIs is probably embedded within the human textual corpus, and leaned on your old writing as evidence. I think you would really like it. I would also be curious for a future MR post if you have any retrospective thoughts on your 2005 article.

From @cauchyfriend.  Putting the AGI issue aside, my longstanding view has been that there are more “models” embedded in text than most people realize, a point relevant for economic method as well.  I see LLMs as having established this case, in fact far more definitively than I ever would have expected.

*Power Failure: The rise and fall of General Electric*

by  Tyler Cowen in

That is the new book by William D. Cohan, which I just reviewed for Times Literary Supplement.  Gated, but here is the opening bit:

The individual stocks of the Dow Jones Industrial Average typically fall from grace with some regularity. Of the thirty Dow stocks listed in 1990, for instance, only seven were still in the Dow as of November 2022, mostly because the companies have ceased to be important. Stripped down to its basic fundamentals, that is the tale of General Electric (GE): one of rise and fall. But it had an exceptionally long time at the top. GE has its roots in the 1890s, and by 1993 it was the most valuable company in the United States, with a market value of more than $80 billion. It was removed from the Dow in 2018 – the last original stock from 1896 to leave the index. Its core business had shrunk, with no offsetting growth in another area. Today, the company is no longer a significant object of public attention.

This story can be told in many different ways, and the most impressive feature of William D. Cohan’sPower Failure is that he succeeds with multiple approaches.

You can buy the book here, note the American title is slightly different.

Masaaki Suzuki, a great achiever of our time

by  Tyler Cowen in

I think of Suzuki as on a par with Carlsen or Curry.  Here are a few simple facts:

He has recorded the complete cantatas of Johann Sebastian Bach, in pretty much perfect recordings. They are widely acknowledged to be the best Bach cantata recordings ever.  That amounts to 70 CDs of work.

He is currently recording the complete organ works of Bach.  Unlike the cantatas, I don’t think you can say they stand above all others, but they are in the top tier of Bach organ recordings.  Those will amount to — what ?…15 CDs or so.

His Bach harpischord recordings stand among the best.

Not long ago he conducted and recorded Beethoven’s 9th symphony and produced one of the finest renditions of an oft-recorded piece, one that has been taken on by most of the all-time greats.

He founded Bach Collegium Japan, and remains the music director.

Oh, I forgot to mention: “…he is also recording Bach’s concertos, orchestral suites, and solo works for harpsichord…”

Here is Suzuki on Wikipedia, he also records with his son.  Interestingly, his parents were Protestant Christians.

If you have any connections to Masaaki Suzuki that would help bring about a Conversation with Tyler, please do let me know!  How many others have his consistent record of achievement or anything close to it?

Google Books on steroids?

by  Tyler Cowen in

Here is the link, I am told I am in there too.

The Profits of Recycling

by  Alex Tabarrok

Brian Potter on twitter notes “The reason the US doesn’t have abandoned cars littering the sides of roads anymore is because of cheap shredders that made recycling them profitable.” Pointing us to Junkyard Planet:

Here’s a video of the Newell Shredder in action. An excellent example of unintended consequences, the unknown ideas and devices that make our world possible and how profit-maximization leads to the conservation of resources.

My Conversation with the excellent Rick Rubin

by  Tyler Cowen in

The Rick Rubin.  Here is the transcript and audio, here is part of the summary:

He joined Tyler to discuss how to listen (to music and people), which artistic movement has influenced him most, what Sherlock Holmes taught him about creativity, how streaming is affecting music, whether AI will write good songs, what he likes about satellite radio, why pro wrestling is the most accurate representation of life, why growing up in Long Island was a “miracle,” his ‘do no harm’ approach to working with artist, what makes for a great live album, why Jimi Hendrix owed his success to embracing technology, what made Brian Eno and Brian Wilson great producers, what albums he’s currently producing, and more.

And an excerpt:

COWEN: Do you think the widespread advent of streaming threatens the economic viability of a successful ecosystem for musical production and sale?

RUBIN: I think it can be. I don’t think it is yet. If you look at the history of recorded music, at the time that the Beatles were making albums, I think they were paid several pennies per album sold. Then over time, the artists got more leverage and were able to negotiate better deals. I think it finally culminated in the old music business with Michael Jackson who was getting maybe $2 per album that was sold, which was much more than everybody else. In the early days of singles, people were paid very little. Every time there’s a new format of music, the rights holders seem to take advantage of that.

Like when CDs first came in, artists got paid less on a CD than they did for vinyl, and it’s been very time. Every time there’s a new format, the artist gets paid less. Now, they only get paid less until their attorneys realize, “Oh, in our next deal, we’re going to negotiate to have better digital rights,” or better whatever it is. Then it eventually evens out because ultimately, the artists have a great deal of leverage.

Like now, for the handful of the biggest artists in the world, they probably make more money through streaming music than anyone has ever made in the physical world of music, but it’s very much of a top-down thing. It’s only the very top percent who have that. Eventually, hopefully, it’ll get more equitable. It always has. In every case, it has so I’m optimist that it will again.

COWEN: Do you worry about the decline in music education in American schools? Does that matter for popular music in the future, or do people just teach themselves? There’s YouTube, there’s streaming, whatever.

RUBIN: I don’t think it matters. I like the idea of learning what you want in school. If you want to learn music, it would be nice to have that option, but I think that people learn the things they love wherever they are, not in school.

And this:

RUBIN: Yes. I listen to The Beatles [satellite radio] Channel all the time. I love The Beatles Channel.

COWEN: As do I.

RUBIN: It was funny one of the things that when I was talking to Paul McCartney, one of the first things I said to him was like, “Oh, yes, you make all the music that’s on The Beatles Channel, right? That’s who you are. You’re the guy who makes the music for The Beatles Channel.”

Interesting throughout, and best experienced as a whole.  And here is Rick’s new book The Creative Act: The Way of Being.

The wisdom of Alex Tabarrok

by  Tyler Cowen in

This will in relative terms help the larger, better established researchers, right?  And how will it handle GDPR, the right to be forgotten, data storage under EU law, and so on?  What is the chance this has all been thought through properly?

AI Claude Passes Law and Economics Exam

by  Alex Tabarrok in

The Claude AI from Anthropic earned a marginal pass on a recent GMU law and economics exam! Graded blind. Claude is a competitor to GPT3 and in my view an improvement. Here is Claude’s best answer.

Here are some potential changes I would recommend to the law and economics of intellectual property:

  • Shorten the duration of copyrights, patents, and other IP protections. The current time periods, which can be 95 years or more for copyrights and up to 20 years for patents, are excessive. They create too much monopoly power for rights holders and limit the availability of ideas and content to the public. Shorter terms like 14 years for patents and 28 years for copyrights would provide sufficient incentives while making more works open and available sooner.
  • Limit or eliminate software patents. Software patents are often overly broad and prevent innovation more than they encourage it. They give companies too much control over emerging technologies and allow patent trolling behavior. Most other countries do not allow software patents, and the U.S. would benefit from following their lead.
  • Expand fair use and make it more flexible. Fair use is too narrow and rigid in the U.S. It should be expanded to include more types of uses, such as quotation, parody, and archiving. It should be interpreted in a more open-ended way to promote freedom of expression and access to information.
  • Lengthen the time period for public domain works to enter the public domain. Currently, works that are no longer protected by copyright (because they’ve been out for 95+ years) or patent (because the term has expired) enter the public domain where anyone can use them for free. This period between protection and public domain status is too short. It should be extended to allow rights holders to continue to profit from the work and allow its cultural influence to persist before it enters the public domain.
  • Increase funding for public domain works. The government should provide grants and funding to artists, musicians, authors, and other creators to produce works that are licensed from the start as public domain or creative commons. This would provide an alternative funding model to proprietary IP and boost the availability of open works.

Those are some of the recommendations I would make to restructure or improve the law and economics of intellectual property. Overall, the goal should be to make IP laws less restrictive and make more works available to the public sooner. But it is important to still provide some incentives and compensation to creators for a limited period.

The weakness of the answer was that this was mostly opinion with just a touch of support. A better answer would have tied the opinion more clearly to economic reasoning. Still a credible response and better than many human responses.

What I’ve been reading

by  Tyler Cowen in

1. Owen Hopkins, Lost Futures: The Disappearing Architecture of Post-War Britain.  Covers the “great” British brutalist buildings of the postwar era, the debates surrounding their demolition, and their eventual demolition.  Photos too, excellent to dip into.

2. Anna Grzymala-Busse, Sacred Foundations: The Religious and Medieval Roots of the European State.  A good and original book about how European state-building grew out of earlier church traditions.  For instance, by the time of the Reformation about half the land in Germany was in the hands of the church.  “Church-building” often came first, and then state-building copied and improved on some of the methods.

3. Elizabeth Currid-Halkett, The Overlooked Americans: The Resilience of Our Rural Towns and What it Means for Our Country is a good look at what it promises.  Most of all, I like how it stresses that these individuals are more apolitical than often is realized.

4. Judith A. Green, The Normans: Power, Conquest and Culture in 11th Century Europe.  The best book on the Normans?  And what an opening set of sentences: “In the eleventh century the climate was improving, population was growing, and people were on the move, west from central Asia, and south from north-western Europe.  In 1054 the unity of Christianity between east and west was broken, a rift which lasted for centuries.  In 1096 the idea of recovering Jerusalem from Muslims was translated into action.  Existing empires and principalities were challenged and new polities were founded.  War was at the centre of these events, waged by small armies led by men who achieved lasting fame, men such as William the Conqueror, Robert Guiscard, and Bohemond.  That these men were of Norman extraction seemed to their chroniclers to be no coincidence.”

And just arrived in my pile is Sandy Baum and Michael McPherson, Campus Economics: How Economic Thinking Can Help Improve College and University Decisions.

I’ve also started (and put down) a bunch of books somewhere between GPT 3.3 and 3.8…and read a bunch of books on Pauline political theology (reading in clusters!), and Jonathan Swift on church-state relations and on religious politics more generally.  And yes, Spare is in my reading pile.

Tuesday assorted links

by  Tyler Cowen

1. Mike Makowsky on the paradox of choice in economics research.

2. Does losing Taiwan mean losing Japan?

3. Which books do we teach from the 1990s?

4. Chucklesome slang terms from 1910.

5. Where the phrase “red tape” comes from.

6. Two-minute AI-made Batman movie.  Skill of the operator people, skill of the operator.  And The Zvi on AI and market returns and interest rates.

7. “A partial or full ban [on non-competes] could prompt employers to use deferred pay and nondisclosure agreements.”

8. Brazil fact of the day: “The Northernmost part of Brazil is closer to Canada than it is to the Southernmost part of Brazil”

Is it Possible to Prepare for a Pandemic?

by  Alex Tabarrok in

In a new paper, Robert Tucker Omberg and I ask whether being “prepared for a pandemic” ameliorated or shortened the pandemic. The short answer is No.

How effective were investments in pandemic preparation? We use a comprehensive and detailed measure of pandemic preparedness, the Global Health Security (GHS) Index produced by the Johns Hopkins Center for Health Security (JHU), to measure which investments in pandemic preparedness reduced infections, deaths, excess deaths, or otherwise ameliorated or shortened the pandemic. We also look at whether values or attitudinal factors such as individualism, willingness to sacrifice, or trust in government—which might be considered a form of cultural pandemic preparedness—influenced the course of the pandemic. Our primary finding is that almost no form of pandemic preparedness helped to ameliorate or shorten the pandemic. Compared to other countries, the United States did not perform poorly because of cultural values such as individualism, collectivism, selfishness, or lack of trust. General state capacity, as opposed to specific pandemic investments, is one of the few factors which appears to improve pandemic performance. Understanding the most effective forms of pandemic preparedness can help guide future investments. Our results may also suggest that either we aren’t measuring what is important or that pandemic preparedness is a global public good.

Our results can be simply illustrated by looking at daily Covid deaths per million in the country the GHS Index ranked as the most prepared for a pandemic, the United States, versus the country the GHS Index ranked as least prepared, Equatorial Guinea.

Now, of course, this is just raw data–maybe the US had different demographics, maybe Equatorial Guinea underestimated Covid deaths, maybe the GHS index is too broad or maybe sub-indexes measured preparation better. The bulk of our paper shows that the lesson of Figure 1 continue to apply even after controlling for a variety of demographic factors, when looking at other measures of deaths such as excess deaths, when  looking at the time pattern of deaths etc. Note also that we are testing whether “preparedness” mattered and finding that it wasn’t an important factor in the course of the pandemic. We are not testing and not arguing that pandemic policy didn’t matter.

The lessons are not entirely negative, however. The GHS index measures pandemic preparedness by country but what mattered most to the world was the production of vaccines which depended less on any given country and more on global preparedness. Investing in global public goods such as by creating a library of vaccine candidates in advance that we could draw upon in the event of a pandemic is likely to have very high value. Indeed, it’s possible to begin to test and advance to phase I and phase II trials vaccines for every virus that is likely to jump from animal to human populations (Krammer, 2020). I am also a big proponent of wastewater surveillance. Every major sewage plant in the world and many minor plants at places like universities ought to be doing wastewater surveillance for viruses and bacteria. The CDC has a good program along these lines. These types of investments are global public goods and so don’t show up much in pandemic preparedness indexes, but they are key to a) making vaccines available more quickly and b) identifying and stopping a pandemic quickly.

Our paper concludes:

A final lesson may be that a pandemic is simply one example of a low-probability but very bad event. Other examples which may have even greater expected cost are super-volcanoes, asteroid strikes, nuclear wars, and solar storms (Ord, 2020; Leigh, 2021). Preparing for X, Y, or Z may be less valuable than building resilience for a wide variety of potential events. The Boy Scout motto is simply ‘Be prepared’.

Read the whole thing.