BRIEFLY NOTED: For 2023-05-15 Mo
The "Young Lady's Illustrated Primer" is still far off; mushroom coffee; our political reporters work very hard to remain underbriefed; Guerrieri & al. on optimal inflation, Törnberg on polarizatio...
MUST-READ: Can ChatBots Make Education a Capital-Intensive Sector?:
First of all: it is not “artificial intelligence”—it is, rather, ChatBots and ImageBots.
Baumol cost-disease routed-around?:
Barry Eichengreen: Winners and Losers in the AI Arms Race: ‘More than anything, however, economic development depends on human development – that is, on the accumulation of human capital. Where developing countries lack the resources, financial and otherwise, to increase significantly their spending on traditional modes of education, AI holds out hope for providing what is missing. It can be used to design individualized learning assistants capable of providing personalized instruction to students in settings where teachers are in short supply. When it comes to economic development, a bit of additional literacy and numeracy can go a long way. Throughout history, technological change has created both winners and losers. There is no reason why AI, like previous technologies, shouldn’t produce more of the former than the latter…
My view:
We are still very far from the "Young Lady's Illustrated Primer" of Neal Stevenson's "The Diamond Age".
We are much, much closer to the gamification of education.
And we are closer still to mammoth numbers of tireless, extremely speedy research assistants who have the average intelligence of the Internet.
The big question for economic development thus seems to me to be this: will these tools once again boost the range of the technoélite, or will they allow the much less educated masses to borrow and access a good deal of how the élite interface with the world?
Here I want to go and find myself optimistic with respect to the growth of wealth and pessimistic with respect to its distribution. Yes, Chat- and Image-Bots are, potentially, a cyberage General Purpose Technology of the same order of magnitude as the mainframe, the PC, the internet, or (shudder) always-connected social media. Each of those utterly transformed and revolutionized about a quarter of the knowledge sector. ChatBots & c. will do the same again. Thus optimism about growth.
My pessimism about distribution comes from my perhaps focusing too much on the concept of metis. People create complicated systems. We then have to figure out how to deal with them. Always, always, it seems that the theoretical knowledge of how these systems are organized and ought to work is fairly easy. But the practical process-knowledge of how to make them work efficiently—that is very, very hard. We can become very good at the process-knowledge component with enough experience, application, and intelligence. But getting those three is never cheap. It requires the expenditure of lots of resources. Hence: to he who hath, yet more is granted. James Scott is quite good on the place of metis in actually working human social systems. (But I would argue that he is largely unaware of how evanescent it turns out to be in a world with a rapidly advancing technological frontier.)
ONE IMAGE: I Am Told This Really Is Legal in Berkeley:
ONE VIDEO: President Joe Biden vs. Reporters Who Are Careful to Remain Underbriefed:
<https://twitter.com/NickButler/status/1656073683415232513>
Very Briefly Noted:
Laura Millan: Spain’s Drought Shows Climate Change Is Here: ‘An unprecedented lack of rainfall and record-high temperatures gripping southwestern Europe for the second straight year had been projected by scientists—for 2043. The heat… is already here…
Noah Smith: ‘After almost 2.5 years, I've finally reached my initial goal of having 10,000 paying Substack subscribers. I feel like I can now think of Noahpinion as a "big" Substack. Thanks for reading, everyone…
Martha Wells: Murderbot: ‘There was a big huge deal about it, and Security was all “but what if it takes over the station’s systems and kills everybody” and Pin-Lee told them “if it wanted to do that it would have done it by now,” which in hindsight was probably not the best response…
Henrik Karlsson: Internet: a user manual: ‘Repeatedly block the things that least serve you… free[ing] up time and space where the things I value can grow…
Kris King: ‘Turkish president Erdogan, out of concern that a new president will be elected, demanded that Twitter censor the opposition. Elon Musk complied…
Robin Wigglesworth: Stonks: Thank You AI Hype: ‘Where would US equities be without ChatGPT mania?…
Jordan Schneider: China's Data Crackdown: ‘Think tank open source research spooks the CCP…
Pandoc: Pandoc Users’ Guide…
The Jolly Swagman Podcast: Intellectual Exoskeletons—Andy Matuschak…
Taegan Goddard: ‘Team Biden was “positively giddy”…. And the best part for Democrats? They have [Trump] on video saying all of these things. Said one Biden aide: “It was quite efficient. Weeks worth of damning content in one hour”…
¶s:
Veronica Guerrieri & al. (2021): Monetary Policy in Times of Structural Reallocation: ‘Asymmetric shocks that shift demand from one sector to another… [produce] an endogenous cost-push shock… resulting in inflation optimally exceeding its target despite elevated unemployment…. When labor is mobile between sectors, monetary easing can have the additional benefit of inducing faster reallocation, by producing wage increases in the expanding sector…
Petter Törnberg: How digital media drive affective polarization through partisan sorting: ‘It is not isolation from opposing views that drives polarization but precisely the fact that digital media bring us to interact outside our local bubble. When individuals interact locally, the outcome is a stable plural patchwork of cross-cutting conflicts. By encouraging nonlocal interaction, digital media drive an alignment of conflicts along partisan lines, thus effacing the counterbalancing effects of local heterogeneity. The result is polarization, even if individual interaction leads to convergence. The model thus suggests that digital media polarize through partisan sorting, creating a maelström in which more and more identities, beliefs, and cultural preferences become drawn into an all-encompassing societal division…
Neil Sinhababu: ‘Heidegger defends scapegoating minorities for genocide in Being and Truth, 1933-1934…. It's a “fundamental requirement” “first to make the enemy” “with the goal of total annihilation” “so that Dasein may not lose its edge.”: “An enemy is each and every person who poses an essential threat to the Dasein of the people and its individual members. The enemy does not have to be external, and the external enemy is not even always the more dangerous one. And it can seem as if there were no enemy. Then it is a fundamental requirement to find the enemy, to expose the enemy to the light, or even first to make the enemy…. The struggle is all the fiercer and harder and tougher… it is often far more difficult and wearisome to catch sight of the enemy as such, to bring the enemy into the open, to harbor no illusions about the enemy, to keep oneself ready for attack, to cultivate and intensify a constant readiness and to prepare the attack looking far ahead with the goal of total annihilation”…
Itamar Drechsler & al.: Banking on Uninsured Deposits: ‘Banks hedge the interest rate risk of their assets with their deposit franchise: when interest rates rise, the value of the assets falls but the value of the deposit franchise rises. Yet the deposit franchise is only valuable if depositors remain in the bank. This creates run incentives for uninsured depositors. We show that a run equilibrium is absent at low interest rates but appears when rates rise because the deposit franchise comes to dominate the value of the bank…. The bank thus faces a dilemma: it cannot simultaneously hedge its interest rate risk and liquidity risk exposures. The dilemma disappears only if uninsured deposits do not contribute to the deposit franchise (if they have a deposit beta of one)…. An optimal capital requirement… rises with interest rates…
Scott Lemieux: “Trump supporters like Donald Trump” is not actually newsworthy: ‘Get a load of this from Chris Licht…. There is, admittedly, a good deal of purchase in the idea that the media didn’t take Trump seriously as a presidential candidate, which definitely affected coverage of the 2016 election, almost entirely for the worse. But taking from this the “lesson” that “we need to confirm that people who like Donald Trump like Donald Trump” is a way of avoiding reflection about these choices… that can also be quite profitable…
What does "We are still very far from the "Young Lady's Illustrated Primer" of Neal Stevenson's "The Diamond Age" mean?
ChatGPT
"The Young Lady's Illustrated Primer" is a fictional book in Neal Stephenson's novel "The Diamond Age." The book is a highly advanced, interactive learning tool that adapts to the needs and interests of its young reader, providing education, guidance, and entertainment.
When someone says, "We are still very far from the 'Young Lady's Illustrated Primer' of Neal Stephenson's 'The Diamond Age'," they mean that we have not yet achieved the level of advanced technology and sophistication depicted in the book. It suggests that while we have made significant progress in the field of interactive education and personalized learning, we have not yet reached the level of technological advancement that would allow us to create something like the Primer in real life.
Ages ago, I joined the gang at the MIT AI Lab for a screening of Colossus: The Forbin Project about an artificially intelligent computer that took over the world. It was pretty awful as a movie, but even worse in its portrayal of artificial intelligence. At various points, the audience would start shouting "Control-S! Control-S!" That's the ASCII code to stop transmission, something very important back in the day when computers and devices like mechanical teletypes could be swamped by incoming data.
I hadn't thought about this until recently when I went to the Open AI site and asked a few cooking questions. The answers weren't bad, but they were massively verbose. The answer to "How long does it take to drip brew a pot of coffee?" shouldn't be longer than a few words or perhaps a full sentence. Well, it produced a reasonably accurate answer in the form of a sentence, then it went on and on and on. It produced a full paragraph of irrelevant material. Then it started to crank out another paragraph. Then, it kept going. I hit control-S and thought about the time I watched Colussus: The Forbin Project with the gang from the MIT AI Lab. Then I found the stop gibbering button and pressed it. Mercifully, it stopped, unlike the movie.
To be honest, I didn't read beyond the first sentence that more or less answered my question. What little of the additional text I scanned seemed like the stuff one wrote when one had to write five hundred words on why I won't make funny noises in class. If this is a productivity enhancement, our society is doomed. Too much first world productivity comes from producing fancy consultant reports full of this kind of ipsum lorem, empty words to pad out arguments agreeing with the sponsor's proposal. Do we really need more and longer reports of this sort? What are the chances that automating such report's production would free up talent for useful work? We all know the answer to that.