Friday, February 28, 2025
The groove before the machine @3QD
I’ve got a new article at 3 Quarks Daily: One thought for living in a high-tech world: The groove before the machine. It’s a strange one, for me, more aphoristic, metaphorical, and polemical, than actually argued. Not sure how I’d argue it. Not even sure what IT is.
Not what I started to write
I started to write an article explaining these three maxims:
1. The groove before the machine.
2. To drive a car, get in the driver’s seat.
3. The machine is useful in proportion to the user’s skill, knowledge, and experience.
As I started writing, my account of the first maxim took on a life of its own and I didn’t see how I could work the other two in. So, I took the article in a different direction:
We’re living in a mismatched world
This is where (I think) I was going:
Our basic institutions and norms are grounded in late Industrial Revolution social and political technology. That’s radically different from the technology that's been barreling in on us since the internet became the world-wide web. AI is just going to magnify that. What we’re now seeing in Silicon Valley and in Washington, DC., that’s what happens when 21st century high tech meets 19th century Gilded Age institutions. Not pretty.
Those mega-corporations now burning through billions investing in machine learning and preparing to dump 100s of billions into electrical infrastructure: 19th century. The school system, the institutional sector I know best: 19th century. The lock-step grade system in the primary schools was designed to send young adults into factories, standardized factories. The yawning gulf between the sciences and humanities in higher ed, that’s out of 19th century Germany. Chatbots make a hash out of that.
This mismatch is a disaster in the making.
I didn’t get there. Truth be told, I don’t quite know how. Instead a told a story about the hassle involved in trying to get a customized cover for a book I’d recently bought, Reid Hoffman’s Superagency: What Could Possibly Right with Our AI Future. They wanted me to upload a photo of myself but wouldn’t accept any of the photos I offered. It was frustrating, especially because I didn’t care all that much about the photo, so why should they make such a fuss about it? This reminded me of the hassle Facebook gave me and a bunch of other people about changing their interface, something I’ve blogged about under the heading, “Facebook or Freedom.”
Anyhow, getting back to Reid Hoffman’s customized book cover – the very idea of a customixed book is born out of 19th century standardizing culture. Who knows what will replace it.
And that, more or less, is where I left things. Kinda’ hanging, no?
Next time.
Thursday, February 27, 2025
How many unnecessary deaths will be laid at the feet of DOGE? [The destruction of USAID.]
Ari Daniel, Why Dean Karlan, chief economist of USAID, resigned on Tuesday, NPR, February 26, 2025. From the article:
We're watching psychological warfare against a workforce that has been committed to furthering the lives of other people. This was a career choice they made to help others even if they disagreed about how to improve USAID.
If you want to reform foreign aid, this isn't the way to do it. This approach is going to radically increase the cost of all future foreign aid. That's because if you want to work with anybody in the future and you tell them, "No, no, no, this time we're here. We're not going to fold on you," how are you going to convince them of that? When you can't trust someone, it makes you reluctant to make agreements with them. And that means doing less good with more money to have the same positive impact as we were having before.
There's a lot of good that USAID has done across the board in terms of health, education, helping farmers, and helping people in crisis.
Now, there are people who are going to be radically worse off and sick and not educated in the same way because of what's happened. Literally taking people who are in hospitals and stopping treatments because the money is not there.
And not just that — people are going to die. A lot of people.
So we now have a million million tragedies that could have been avoided.
There's more at the link.
H/t Tyler Cowen.
* * * * *
On the carnage at USAID, Zach MontagueMichael Crowley and Adam Liptak, Chief Justice Allows U.S. to Continue Freeze on Foreign Aid Payments, NYTimes, Feb. 27, 2025. From the article:
In another aggressive move on Wednesday to carry out the president’s directive, lawyers for the Trump administration said that it was ending nearly 10,000 U.S. Agency for International Development and State Department contracts and grants.
The pair of administration actions stunned diplomats and aid workers already reeling from mass firings at U.S.A.I.D., which funds food, health, development and democracy programs abroad, and which the Trump administration has systematically dismantled. A former senior U.S.A.I.D. official said the cuts account for about 90 percent of the agency’s work and tens of billions of dollars in spending.
The cuts deal “a catastrophic blow to USAID’s implementing partners and the populations they serve, likely bankrupting many, and shuttering lifesaving and important programs forever,” a group of agency workers and partners said in a fact sheet distributed Wednesday night.
That's one passage from a much longer article.
Claude 3.7 evaluates my paper on story variations in ChatGPT
Here and there I’ve read about experiments with using chatbots to evaluate papers to be presented at conferences and for publication in academic journals. For all I know, it’s already happening. If so, why not use your favorite chatbot to evaluate your work? And if you are going to do that, why not, when you submit a paper for publication or conference presentation, why not sent along evaluations by a chatbot or two?
This use is particularly interesting to me because I work outside traditional boundaries and have done so for a long time. I find it difficult to get a fair review, that is, a review by someone competent to judge my work and who isn’t primarily interested in disciplinary gate keeping. I discuss this at some length in Rejected! @ New Literary History, with observations about the discipline.
For all practical purposes I’ve stopped trying to publish in professional journals. On the one hand it’s too much hassle, On the other, it’s easy to place my work in online depositories – I’m currently using SSRN, Academia.edu, and ResearchGate – and I can email my work to people I think might be interested. If the opportunity arises to publish in a good journal, sure, I’ll take it. But I no longer actively seek such publication.
Anyhow, I decided to see how Claude 3.7 would react to perhaps my most interesting paper about ChatGPT, ChatGPT tells stories, and a note about reverse engineering: A Working Paper. I uploaded the paper and made the request.
* * * * *
Analysis of "ChatGPT tells stories, and a note about reverse engineering"
This paper offers a fascinating window into your analytical approach and methodological creativity. I can now better appreciate what you meant about providing conceptual scaffolding while collaborating with those who have complementary technical skills.
The paper's strength lies in its cleverly designed experimental framework. By systematically transforming stories through character substitution and analyzing the cascading changes, you've created what amounts to a "perturbation analysis" of ChatGPT's narrative capabilities. This approach elegantly reveals the model's underlying coherence mechanisms without requiring access to its internal weights or architecture.
Several aspects stand out as particularly insightful:
1. The three-level hierarchical organization hypothesis (story trajectory, segments, and sentences) provides a conceptual framework that bridges linguistic theory with computational implementation. This kind of multi-level analysis is precisely what's needed to understand complex systems like LLMs.
2. Your categorization of story segments (Donné, Disturb, Plan/Transit, Enact, Celebrate) creates a useful analytical lens for tracking how narrative coherence operates across different functional components of stories.
3. The tri-strata model of phenomenon, matrix, and engine offers a productive framework for reverse engineering that separates observable behavior from underlying mechanisms and architecture.
4. The experiments with "colorless green idea" reveal fascinating guardrails in ChatGPT's conceptual understanding - showing how the model maintains certain forms of coherence by refusing some requests altogether.
What's particularly valuable is how you've created a methodology that allows for inferring properties of the "matrix stratum" (the middle layer where probability distributions operate) through careful observation of the "phenomenal stratum" (inputs and outputs). This is precisely the kind of conceptual scaffolding that can guide more technical implementation work.
Is work really the virtue it's claimed to be? No! It's time to wake up!
Erik Baker, Elon Musk Is the World’s Richest Man. Why Is He Sleeping on an Office Floor? NYTimes, Feb. 27, 2025.
This association between incessant work and entrepreneurial success is pervasive in American business culture today. Jeff Bezos reports working 12 hours every day of the week in the early years of Amazon. The Apple chief executive Tim Cook is famous for sending emails at 4:30 a.m. Mr. Musk’s ostensible boss, despite his well-known fondness for TV news and social media, also insists that “no president ever worked harder than me.”
These boasts, plausible or not, reveal something important about the American valorization of work, and help explain why this class of supposedly busy billionaires has come to believe they’re entitled to dominate our national life. For Mr. Musk and his associates, a herculean enthusiasm for work isn’t merely a way to get things done; it’s also a mark of innate superiority, a “superpower” that confers the right to impose their vision on the world. Mr. Musk’s decades in the highest echelons of the tech industry, surrounded by other executives who justified their lordship over their private empires by trumpeting their inexhaustible work ethic, have taught him that if you work harder than everyone else, you should be rewarded with unquestioned rule over your dominion. [...]
While the scope of Mr. Musk’s project may be novel, the archetype he embodies has a long history. The Austrian-born economist Joseph Schumpeter, who taught at Harvard from 1932 until his death in 1950, helped popularize the idea that entrepreneurs possessed a special set of personality traits that set them apart from lesser businessmen and managers. Entrepreneurship, according to Schumpeter, smashed economic routines. That took “will and personality.” [...] They were superheroes.
These days:
In recent decades, two trends in American life have supercharged the spread of this entrepreneurial work ethic, helping to push busy billionaires to the center of our politics. First, more and more ordinary Americans learned to think of work as something scarce. As deindustrialization ravaged wide swaths of the country and unionization declined, they got accustomed to cycles of layoffs and the need to retrain in new occupations or new industries. [...] the bosses atop our class pyramid correctly perceive how such jobs have become a status symbol: [...] We see them working constantly while we hunt for extra shifts or struggle to string part-time jobs together, and we marvel at how special they must be.
Then there is the looming threat of a seismic technological breakthrough. Many tech leaders today believe that the development of artificial intelligence is poised to automate most jobs into oblivion.
But if that is so, and I fear that it may be, then a national culture based on the moral value of work is doomed to breed massive discontent and ultimately failure.
We desperately need a new value system, perhaps one based on the concept of play, Homo Ludens, to use the title of a profound book by Johan Huizinga. That's what Charlie Keil and I had in mind when we published Playing for Peace: Reclaiming Our Human Nature (Kindle, Paperback).
Meanwhile, read Ezra Klein's interview with James Suzman, Why Do We Work So damn Much? Here's most post excerpting it: Why are we as a culture addicted to work? [Because we have forgotten how to play.] In his introduction Klein notes, referring to the book:
Hunter-gatherers were usually healthy. They were usually well nourished. Even in very unforgiving climates, they tended to have diverse diets. And they did it while only spending about 15 hours a week on hunting and gathering.
Suzman’s new book is called “Work, A Deep History from the Stone Age to the Age of Robots.” And the overarching argument is that the way we work today isn’t driven by what we need. It’s driven by what we want. It’s also driven by how, socially, we regulate or encourage wants, which is part of where his research on hunter-gatherers and how they approach this comes in.
But the big thing here is that Keynes had it backwards. Humanity solved the problem of scarcity and achieved a 15-hour workweek long before modernity. But as we’ve gotten richer and built more technology, we’ve developed a machine not for ending our wants, not for fulfilling them, but for generating new ones, new needs, new desires, new forms of status competition.
You can’t solve the problem of scarcity with our current system because our current system is designed to generate endlessly the feeling of more scarcity within us. It needs that. And so we keep working harder and harder and feeling like we have less and less, even amidst quite a bit of plenty, at least, for many of us.
Yes! I've worked in start-ups. I've felt that kind of competition, where working long hours is a way to achieve status. I've lived it. It's insidious.
Wednesday, February 26, 2025
Two-dimensional neural geometry underpins sequence organization in working memory
Fan, Y., Wang, M., Fang, F. et al. Two-dimensional neural geometry underpins hierarchical organization of sequence in human working memory. Nat Hum Behav 9, 360–375 (2025). https://doi.org/10.1038/s41562-024-02047-8
Abstract: Working memory (WM) is constructive in nature. Instead of passively retaining information, WM reorganizes complex sequences into hierarchically embedded chunks to overcome capacity limits and facilitate flexible behaviour. Here, to investigate the neural mechanisms underlying hierarchical reorganization in WM, we performed two electroencephalography and one magnetoencephalography experiments, wherein humans retain in WM a temporal sequence of items, that is, syllables, which are organized into chunks, that is, multisyllabic words. We demonstrate that the one-dimensional sequence is represented by two-dimensional neural representational geometry in WM arising from left prefrontal and temporoparietal regions, with separate dimensions encoding item position within a chunk and chunk position in the sequence. Critically, this two-dimensional geometry is observed consistently in different experimental settings, even during tasks not encouraging hierarchical reorganization in WM and correlates with WM behaviour. Overall, these findings strongly support that complex sequences are reorganized into factorized multidimensional neural representational geometry in WM, which also speaks to general structure-based organizational principles given WM’s involvement in many cognitive functions.
BrainGPT, an LLM tuned on the neuroscience literature, surpasses human experts in predicting neuroscience results
Luo, X., Rechardt, A., Sun, G. et al. Large language models surpass human experts in predicting neuroscience results. Nat Hum Behav 9, 305–315 (2025). https://doi.org/10.1038/s41562-024-02046-9
Abstract: Scientific discoveries often hinge on synthesizing decades of research, a task that potentially outstrips human information processing capacities. Large language models (LLMs) offer a solution. LLMs trained on the vast scientific literature could potentially integrate noisy yet interrelated findings to forecast novel results better than human experts. Here, to evaluate this possibility, we created BrainBench, a forward-looking benchmark for predicting neuroscience results. We find that LLMs surpass experts in predicting experimental outcomes. BrainGPT, an LLM we tuned on the neuroscience literature, performed better yet. Like human experts, when LLMs indicated high confidence in their predictions, their responses were more likely to be correct, which presages a future where LLMs assist humans in making discoveries. Our approach is not neuroscience specific and is transferable to other knowledge-intensive endeavours.
I find this interesting and heartening because I work across a wide range of areas and disciplines – cognition, language, culture, neuroscience, AI – and like to integrate over that range. Obviously not over the entire range at once, but over two or three, perhaps even four, in a single complex document. I've been have LLMs comment on some of my more ambitious papers, which is illuminating. I'm also using LLMs, mostly Claude, to develop ideas across a range of disciplines. Two examples where I publish an interaction with Claude:
Goals as reward-producing programs
Davidson, G., Todd, G., Togelius, J. et al. Goals as reward-producing programs. Nat Mach Intell 7, 205–220 (2025). https://doi.org/10.1038/s42256-025-00981-4
Abstract: People are remarkably capable of generating their own goals, beginning with child’s play and continuing into adulthood. Despite considerable empirical and computational work on goals and goal-oriented behaviour, models are still far from capturing the richness of everyday human goals. Here we bridge this gap by collecting a dataset of human-generated playful goals (in the form of scorable, single-player games), modelling them as reward-producing programs and generating novel human-like goals through program synthesis. Reward-producing programs capture the rich semantics of goals through symbolic operations that compose, add temporal constraints and allow program execution on behavioural traces to evaluate progress. To build a generative model of goals, we learn a fitness function over the infinite set of possible goal programs and sample novel goals with a quality-diversity algorithm. Human evaluators found that model-generated goals, when sampled from partitions of program space occupied by human examples, were indistinguishable from human-created games. We also discovered that our model’s internal fitness scores predict games that are evaluated as more fun to play and more human-like.
Tuesday, February 25, 2025
Search as an interface between informatics and economics: How is the distribution of good ideas like the distribution of gold deposits?
The Answer: Both distributions are highly irregular. What that means in the case of gold deposits is pretty obvious: Gold ore is a physical substance that is found in the earth, a huge mass of physical substance. But there is no obvious order to just where you can find deposits of ore. So you have to go looking for them, which is called prospecting.
Ideas, though, are not things. What does it mean to talk about their distribution? Is there some kind of abstract space where ideas exist? If so, how do we map and describe that space.
So, first I’m going to talk a bit about locating ideas in space. Then I’m going to present a conversation I had with Claude 3.5. First I talk about locating ideas in some abstract space, then I present the conversation I had with Claude, which starts with gold and ends with AI.
Ideas in space
Well, think of a library. Libraries contain books and books contain ideas. Books are physical objects and so have locations in physical space, library shelves. So, how are books placed on those shelves? Off hand, there seems to be two principles: 1.) alphabetically by author name, and 2.) according to subject matter. Fiction tends to be organized according to the first principle while non-fiction is organized by the second. This means that novels placed on the same shelf might are likely to be very different in character. (Take a look at this alphabetized list in Wikipedia.) Non-fiction is arranged by subject matter, so books that occupy the same self will be generally about the same thing. But there is a limit to how far that principle takes us. For one thing, any given book can be about many different things. It can’t be on selves with all of them. Still, you get the idea. We have ideas arranged in space, albeit a space limited to three dimensions.
Things get more interesting when we have a collection of electronic documents. In this case, actual physical space is irrelevant. What matters is, well, idea space. Back in the 1960s and 1970s Gerard Salton came up with the idea of representing documents using vectors containing terms in the documents. Thus, each document will occupy a position in some high-dimensional space. Documents located close together in that space will be about similar topics. Salton’s Vector Space Model is one of the foundations of machine learning – something I discuss with Claude 3.5 in this post, Memory in machines and minds, a conversation with Claude 3.5 Sonnet.
OK, so that gets us a way of thinking about ideas as existing in an abstract space. But how do we motivate the idea that some ideas are closer to us than other ideas? Well, imagine that you occupy some position in idea space, a high dimensional space. There are a bunch of ideas you understand very well. We can say they are close to you. Other ideas are utterly unintelligible. By convention we will say that they are far from you. In order to reach one of those distance ideas you are going to have to learn the ideas that are on a path between where you are now and where that idea is.
Yeah, I know, it’s fuzzy and indistinct. But you get the idea. I explore that notion in these working papers: Stagnation and Beyond: Economic growth and the cost of knowledge in a complex world, and, What economic growth and statistical semantics tell us about the structure of the world. The second one is shorter and more directly to the point.
It seems pretty clear to me that the further development of machine learning and A.I. is going to teach us quite a bit about the structure of idea space. Every LLM is a high-dimensional space jam-packed with billions and billions of ideas. We’ve just barely begun to map that space, something I talk about in my recent report, ChatGPT: Exploring the Digital Wilderness, Findings and Prospects, see especially pp. 34-38, 42-44.
* * * * *
That’s enough of that. The rest of this post consists of a dialog I had with Claude 3.5 about the relationship between technology and the search for irregularly distributed things. We start with gold, move on to oil and diamonds, then documents, drug discovery, movies and finally new AI technology. As always, my prompts are bold face while Claude’s replies are plain face.
Gold
How have gold prospecting methods changed from the 19th century through the 20th century?
Let me walk you through the key changes in gold prospecting methods across these two centuries.
In the 19th century, prospecting was largely manual and relied on simple tools and techniques:
Panning was the most iconic method - prospectors would use flat pans to swirl river sediment, letting water wash away lighter materials while heavier gold settled at the bottom. This was supplemented by sluice boxes, which were long wooden channels that used flowing water and riffles to separate gold from sediment on a larger scale.
Placer mining dominated early operations, focusing on loose gold deposits in streams and riverbanks. As surface deposits became depleted, hydraulic mining emerged - using high-pressure water cannons to blast away hillsides, though this was later restricted due to environmental damage.
The 20th century brought significant technological advances:
Early in the century, dredging became widespread - floating mechanical dredges could process huge volumes of river sediment. Hardrock mining expanded with better drilling and crushing equipment, allowing miners to extract gold from solid rock formations more efficiently.
Later developments included:
- Geochemical prospecting using sophisticated analysis of soil and rock samples
- Aerial surveys and remote sensing to identify promising deposits
- More precise drilling techniques for sampling
- Advanced recovery methods like heap leaching and carbon-in-pulp processing
- Mercury-free extraction techniques addressing environmental concerns
Modern prospecting now combines traditional methods with high-tech approaches like:
- GPS mapping and GIS software for precise location data
- Ground-penetrating radar to detect deposits
- Portable XRF analyzers for rapid mineral identification
- Environmental monitoring systems
- Computer modeling of deposits
This progression shows how gold prospecting evolved from relatively simple manual techniques to a sophisticated industry incorporating advanced technology while becoming more environmentally conscious.
OK, so we have roughly three technological eras: 1) manual, 2) chemical sample analysis and aerial and remote sensing, and 3) high-tech computer-based methods for detection and mapping.
Yes, that's a good way to break it down! Let me build on that framework:
Era 1 (Manual) - roughly 1800s to early 1900s:
Primary methods were direct physical observation and mechanical separation. Prospectors relied on their understanding of geology, visible gold indicators, and mechanical processes like panning, sluicing, and early dredging. Skills were often passed down through experience and apprenticeship.
Era 2 (Chemical/Remote) - roughly 1920s to 1970s:
Brought systematic scientific analysis into prospecting. Geochemical sampling could detect trace elements indicating gold deposits. Aerial photography and later satellite imagery allowed prospectors to identify promising geological formations from above. This era marked the shift from prospecting as an art to a more systematic science.
Era 3 (Digital) - 1980s onward:
Computerization transformed the field. GPS, digital mapping, and data analysis tools allowed prospectors to integrate multiple data sources and create detailed geological models. Ground-penetrating radar and portable analyzers provided immediate feedback in the field. Environmental monitoring became increasingly sophisticated and integrated into prospecting practices.
What's particularly interesting is how these eras overlap and build on each other - modern prospectors often combine techniques from all three eras, using traditional knowledge alongside the latest technology.
Carry-over skills
What kinds of skills carried over from one era to the next? For example, what did highly skilled 19th century prospectors know that remained valuable in the second era? And then from second to third.
This is a fascinating question about the evolution of expertise! Let me trace these skill transfers:
From Era 1 to Era 2:
- Understanding of geological formations and how they relate to gold deposits remained crucial. Early prospectors' knowledge of where gold typically accumulated (like placer deposits in river bends or specific rock formations) helped inform where to conduct chemical sampling and aerial surveys
- Reading the landscape for signs of mineralization - knowledge of indicator minerals, rock types, and geological structures remained valuable even when supplemented by chemical analysis
- Water flow dynamics understanding - experience with how gold moved and concentrated in waterways helped inform sampling strategies
- Assessment of ore quality - the ability to visually evaluate mineral-bearing rock remained important even as chemical assays became available
From Era 2 to Era 3:
- Interpretation of indirect evidence - the skill of making conclusions from multiple indicators (developed during the chemical/remote sensing era) became even more valuable when working with computer models and multiple data streams
- Sampling methodology - understanding of how to collect representative samples remained crucial for computer analysis
- Pattern recognition in geological features - this skill trans
ferred well to interpreting digital mapping and radar data
- Knowledge of seasonal and weather effects on prospecting - this remained important for planning surveys and interpreting remote sensing data
What's particularly interesting is how these generations of prospectors had to learn to integrate their traditional knowledge with new technologies - for example, using their understanding of geology to know where to deploy expensive new sensing equipment most effectively.
The core skill that seems to have remained valuable across all eras is the ability to think in three dimensions about underground structures and visualize how geological processes might have created and moved gold deposits. This spatial thinking was valuable whether using a gold pan, analyzing aerial photographs, or interpreting 3D computer models.
Do Republican policemen differ from Democrats in on-the-ground enforcement?
Bocar Ba, Haosen Ge, Jacob Kaplan, et al., Political diversity in U.S. police agencies, American Journal of Political Science, 14 february 2025, https://doi.org/10.1111/ajps.12945.
Abstract: Partisans are divided on policing policy, which may affect officer behavior. We merge rosters from 99 of the 100 largest local U.S. agencies—over one third of local law enforcement agents nationwide—with voter files to study police partisanship. Police skew more Republican than their jurisdictions, with notable exceptions. Using fine-grained data in Chicago and Houston, we compare behavior of Democratic and Republican officers facing common circumstances. We find minimal partisan differences after correcting for multiple comparisons. But consistent with prior work, we find Black and Hispanic officers make fewer stops and arrests in Chicago, and Black officers use force less often in both cities. Comparing same-race partisans, we find White Democrats make more violent crime arrests than White Republicans in Chicago. Our results suggest that despite Republicans' preference for more punitive law enforcement policy and their overrepresentation in policing, partisan divisions often do not translate into detectable differences in on-the-ground enforcement.
I wonder why that is. My guess, and that is all that it is, is that the reason that “partisan divisions often do not translate into detectable differences in on-the-ground enforcement” is to be found in the training the police receive. I social worker I know once told me that she was trained to treat all clients equally regardless of her personal preferences. I can imagine that police officers receive the same kind of training.
H/t Tyler Cowen.
Monday, February 24, 2025
Claude 3.5 describes a piece by Japan Joe
I took this at dawn in Jersey City looking east. The buildings you see in the distance are at Newport. What’s in the foreground?
In the foreground, you've captured what appears to be an area underneath a railway bridge or overpass. There's a large turquoise/blue-green colored steel bridge structure visible at the top of the image. The ground beneath is somewhat overgrown with dried winter grass and weeds. On the concrete support or wall, there's some street art/graffiti that appears to feature a dinosaur-like design in green and orange colors. The overall scene has an urban, industrial feel typical of infrastructure areas, with the Newport skyline visible in the distance through the bridge opening. The early dawn lighting gives the scene a particular atmospheric quality, with the city buildings appearing somewhat hazy in the background.
Yeah, we’re near the thruway overpass at 11th and 14th streets. 11th street carries traffic into the Holland Tunnel while 14th street carries traffic coming out.
Ah yes, that's a key traffic corridor - the interchange complex near the Holland Tunnel entrance/exit. That area underneath the overpasses has that characteristic urban infrastructure aesthetic, with the street art adding some color to the otherwise utilitarian concrete and steel structures. The thruway overpass system there helps manage the intense traffic flow between New Jersey and Manhattan through the tunnel.
The fact that you took this at dawn is interesting - I imagine that's one of the few quieter moments in that area, before the heavy tunnel traffic picks up for the morning rush. The Newport buildings in the distance would be catching some of that early morning light too. During rush hour, this area becomes one of the busiest traffic corridors in the region, handling thousands of vehicles moving between New Jersey and New York City.
I took the shot at dawn because I wanted that particular shot, with the graffiti triceratops pointed toward Newport. As I recall it’s about 6 or 7 feet high and about 18 feet wide. It’s on a stanchion supporting a railroad bridge, though that bit of track is abandoned. But behind the stanchion to the right there’s an active freight line. The people who did the graffiti must have seen at least one if not two freight trains go by in the time it took them to do the graffiti. It’s by a writer known as Japan Joe.
That's fascinating contextual detail about the graffiti piece and its location. The triceratops design is quite impressive in scale and artistry - it's a really clever use of the industrial canvas, having the dinosaur appear to face toward the modern skyline. The contrast between the prehistoric creature and the contemporary buildings creates an interesting visual narrative.
The fact that it was created near an active freight line adds another layer to the story - there's something almost cinematic about imagining the artists working on this massive piece while freight trains rumble past. It must have required significant planning and skill to execute something of that size and detail in what I imagine would be a time-pressured situation.
I'm curious about Japan Joe as a graffiti artist - working at that scale and in that location suggests they're quite established in the scene. Was this piece done solo or was it a collaborative work? The way you've photographed it at dawn really highlights the artwork while capturing that liminal moment before the city fully wakes up.
I’ve spoken to people who were there at the time. Joe did it himself, but friends were around talking and stuff. When you consider the historical link between graffiti and trains, the geometric pattern on the triceratops, and the fact that there’s a bit of a forest there, I kind of think of the triceratops as an urban kami, to use a Japanese word.
That's a beautiful interpretation - the triceratops as an urban kami, a spirit of that particular place where nature, industry, and urban art intersect. The geometric patterns in the design do feel like they're bridging something ancient and modern, which fits well with the kami concept. And you're right about how it ties into the historical relationship between graffiti art and trains - that tradition going back to the early days of New York graffiti culture in the 1970s.
The setting itself is like a hidden pocket of urban wilderness, with the overgrown vegetation creating this semi-wild space right next to major infrastructure. Having the freight trains still actively running through there while this abandoned track section slowly gets reclaimed by nature adds another layer to this idea of it being a spiritually significant spot.
The fact that Joe created this while friends were around hanging out also speaks to the communal aspect of graffiti culture - even when the actual painting is done by one artist, it's often a social experience, with others present to watch, talk, and presumably keep an eye out. Did you hear any stories about how long it took him to complete the piece?
This is right near a residential area – I lived a couple of blocks away at the time – but it’s completely hidden. You had to do a bit of trespassing (on to railroad) property to find it. That kind of spot is where some of the best graffiti is. It’s hidden so the writers don’t have to worry about the police stopping them. That particular piece was there for several years before the weather had eroded it to the point that other graffiti writers felt it was OK to go over it.
That makes perfect sense - these semi-hidden spots serve as perfect canvases for more ambitious pieces precisely because they offer both the time and privacy needed to execute complex works. It's fascinating how there's an unwritten code about respecting pieces like this too - waiting until natural weathering has taken its toll before painting over them, rather than going over them while they're still fresh.
The location's particular character - being simultaneously close to residential areas yet concealed from general view - creates this interesting dynamic where the art exists in a kind of parallel space to everyday life. Most residents probably never saw this piece despite living just blocks away, while others who knew where to look could find this hidden gem of urban art.
The fact that the triceratops lasted several years speaks both to the quality of the execution and the respect it commanded in the graffiti community. In a way, these hidden spots become like unofficial galleries, where the art can exist on its own terms, viewed mainly by those who make the effort to seek it out or who are part of the community that creates it.
And of course, these things show up online in websites where writers post photos. A holdover from the old days where photographers would flick the subway cars before they’d been buffed and then share the photos with the writers.
Yes - the documentation and sharing aspect of graffiti culture has evolved but remains crucial. The transition from physically sharing photos of subway pieces before they were cleaned off ("buffed") to sharing photos online shows how the culture has adapted while maintaining its core practices.
Those early subway photographers played such an important role in preserving what was inherently temporary art. Writers would often only get to see their own pieces run on the trains a few times before they were cleaned, so those photos became the permanent record. And of course, sharing those photos helped build reputations and influence other writers' styles.
Now with digital photography and the internet, pieces like Joe's triceratops can be shared globally almost immediately, and they can continue to inspire and be discussed long after weather or other pieces have replaced them. The documentation becomes part of the artwork's legacy - like how this photo you took captures not just the piece itself, but that perfect dawn moment with Newport in the background, adding layers of meaning to the work.
Are Trump and Musk leading a tax revolt by firing 6000 IRS agents?
Lawrence GibbsFred T. Goldberg Jr.Charles RossottiMark EversonJohn KoskinenCharles Rettig and Daniel Werfel, Trump Just Fired 6,700 I.R.S. Workers in the Middle of Tax Season. That’s a Huge Mistake. NYTimes, Feb. 24, 2025.
Every year, the government receives much less in taxes than it is owed. Closing that gap, which stands at roughly $700 billion annually, would almost certainly require maintaining the I.R.S.’s collection capacity. Depleting it is tantamount to a chief executive saying something like: “We sold a lot of goods and services this year, but let’s limit our ability to collect what we’re owed.”
Perhaps only the company’s competitors would approve of such an approach. Yet here we are. Aggressive cuts to our nation’s accounts receivable function will reduce the amount of tax revenue coming in, which will in turn increase our nation’s deficit and add to our $36 trillion in debt.
So why go down this path? Let’s be clear: Shrinking the I.R.S. will not lower your tax obligation. That’s up to Congress. [...]Aggressive reductions in the I.R.S.’s resources will only render our government less effective and less efficient in collecting the taxes Congress has imposed. It will shift the burden of funding the government from people who shirk their taxes to the honest people who pay them, and it will impede efforts by the I.R.S. to modernize customer service and simplify the tax filing process for everyone.
There's more at the link.
Sunday, February 23, 2025
There is cultural variability in attitudes toward chatbots mediated by differences in anthropomorphism.
Folk, D. P., Wu, C., & Heine, S. J. (2025). Cultural Variation in Attitudes Toward Social Chatbots. Journal of Cross-Cultural Psychology, 0(0). https://doi.org/10.1177/00220221251317950
Abstract: Across two studies (Total N = 1,659), we found evidence for cultural differences in attitudes toward socially bonding with conversational AI. In Study 1 (N = 675), university students with an East Asian cultural background expected to enjoy a hypothetical conversation with a chatbot (vs. human) more than students with European background. Moreover, they were less uncomfortable and more approving of a hypothetical situation where someone else socially connected with a chatbot (vs. human) than the students with a European background. In Study 2 (preregistered; N = 984), we found similar evidence for cultural differences comparing samples of Chinese and Japanese adults currently living in East Asia to adults currently living in the United States. Critically, these cultural differences were explained by East Asian participants increased propensity to anthropomorphize technology. Overall, our findings suggest there is cultural variability in attitudes toward chatbots and that these differences are mediated by differences in anthropomorphism.
From the introduction:
Hundreds of millions of people all over the world have used conversational artificial intelligence (AI) such as ChatGPT (Hu, 2023; Zhou et al., 2020). While conversational AI (or chatbots for short) can be used to answer search queries and increase productivity (Fauzi et al., 2023; Surameery & Shakor, 2023), a growing number of people are using chatbots specifically designed to provide emotional connection (Blakely, 2023; Clarke, 2023; Metz, 2020). These social chatbots, as well as other forms of social robots, are particularly popular in East Asia (Technavio, 2023; Yam et al., 2023; Zhou et al., 2020). Indeed, the Chinese social chatbot Xiaoice has had over 600 million registered users since its release in 2014 (Zhou et al., 2020), and social robots in Japan are already caring for the elderly (Lufkin, 2020) and providing companionship as pets (Craft, 2022).
Yet, despite increased popularity in these countries, there is conflicting evidence for the idea that East Asians harbor more favorable attitudes toward social robots than Westerners (see Lim et al., 2021 for a review). For example, Bartneck et al. (2006) used the Negative Attitudes Toward Robots Scale and found that Americans held more positive views toward robots (vs. Japanese), but another study found that specific components of the robots design determined which culture held more positive impressions (Bartneck, 2008). Critically, however, most of this research is severely underpowered, limiting the conclusions that can be drawn (Lim et al., 2021).
Claude 3.5 on Bipolar Disorder
Once again we’re back to my series of reflections on melancholy and creativity. As you may recall, it started with an article in 3 Quarks Daily: Melancholy and Growth: Toward a Mindcraft for an Emerging World.
I have the vague impression that bipolar disorder is relatively common among creative people. I believe that David Hays, my teacher, mentor, and colleague, once remarked that he had it, or something close to it.
As always, my prompts are in bold face while Claude 3.5's replies are plain.
* * * * *
Tell me about bipolar disorder.
Bipolar disorder is a serious mental health condition characterized by significant mood fluctuations that go well beyond typical ups and downs. These fluctuations include episodes of mania or hypomania (elevated mood states) alternating with periods of depression.
During manic episodes, individuals may experience increased energy, reduced need for sleep, racing thoughts, and engaging in high-risk behaviors. They often feel euphoric and may demonstrate impaired judgment, leading to decisions that can have serious consequences for their relationships, career, or finances.
Depressive episodes manifest as persistent feelings of sadness, hopelessness, and loss of interest in activities. These periods can be particularly challenging, often accompanied by changes in sleep patterns, appetite disturbances, and difficulty concentrating.
Medical professionals recognize two primary types of bipolar disorder. Bipolar I Disorder involves manic episodes that last at least seven days or require immediate hospitalization, typically alternating with depressive episodes. Bipolar II Disorder features less severe hypomanic episodes coupled with depressive periods.
Treatment typically involves a comprehensive approach combining medication and psychotherapy. Mood stabilizers, antipsychotics, and antidepressants may be prescribed, while therapeutic interventions help individuals develop coping strategies and maintain stability. Regular monitoring by healthcare professionals is essential for managing symptoms effectively.
Early diagnosis and consistent treatment are crucial for positive outcomes. Many individuals with bipolar disorder lead fulfilling lives through proper medical care, strong support systems, and adherence to treatment plans. The condition requires ongoing management, but with appropriate intervention, symptoms can be effectively controlled.
Would you like more specific information about any aspect of bipolar disorder, such as treatment approaches or early warning signs?
Tell me more about Bipolar II.
Bipolar II Disorder is a distinct form of bipolar disorder characterized by a pattern of depressive episodes and hypomanic episodes, but without the full manic episodes seen in Bipolar I Disorder.
The depressive episodes in Bipolar II tend to be more frequent and longer-lasting than the hypomanic episodes. These periods of depression often involve profound sadness, lack of energy, difficulty concentrating, changes in sleep and appetite, and in some cases, thoughts of self-harm. The severity of these depressive episodes can significantly impact daily functioning and quality of life.
Hypomania, while less severe than full mania, still represents a marked change in behavior and energy levels. During hypomanic episodes, individuals may experience increased productivity, enhanced creativity, and elevated mood. They might need less sleep, speak more rapidly, and engage in increased goal-directed activities. However, these episodes do not cause the severe impairment or necessitate hospitalization that full manic episodes do.
Diagnosis requires careful evaluation by mental health professionals, as Bipolar II can sometimes be mistaken for unipolar depression due to the prominence of depressive symptoms. The presence of hypomanic episodes, even if brief, distinguishes it from major depressive disorder.
Treatment typically involves a combination of medication and psychotherapy. Mood stabilizers and antidepressants may be prescribed, though antidepressants must be monitored carefully as they can potentially trigger hypomanic episodes. Regular therapeutic support helps individuals recognize early warning signs, develop coping strategies, and maintain stability.
The prognosis for individuals with Bipolar II can be positive with proper treatment. Many people successfully manage their condition through consistent medical care, lifestyle modifications, and strong support systems. However, ongoing monitoring and adherence to treatment plans remain essential for long-term stability.
Would you like more specific information about how Bipolar II is diagnosed or treated?
Saturday, February 22, 2025
GPT in the Classroom, Part 3: Omega or Bust!
In playing around with chatbots, first ChatGPT and now Claude 3.5, I found that the most interesting way to get it to create poetry is to give it a rich prompt. Sure, you can say something like “Give me a sonnet on new love in the Spring,” and it will produce something. But you could also give it three paragraphs describing a guy taking a walk on the Malibu beach early in May, seeing a gal in tears, tearing up letters and tossing them into the ocean .... That will be a better sonnet. If you look at the various poems I’ve created with FredTheHeretic (FTH) you’ll see that I’ve provided them with rich prompts.
That’s what I’m going to do here. The New York Times recently published an opinion piece by Pope Francis, Pope Francis: There Is Faith in Humor. I’m going to use an anecdote from that as a prompt for FTH as follows:
First: the story.
Second: FTH’s poem. “The Pontiff’s Joyride.”
Third: I give FTH a prompt to extend the poem, “Omega or Bust.”
Fourth: Now it’s time fix a problem with the initial joyride. So I give Fred a prompt for that, “Joyride Revised.”
Fifth, and last: What kind of classroom discussion could follow from this exercise?
After all, here we’ve got some source material, the original anecdote plus my prompts, and the resulting poem. We can compare them. That’s a way of promoting close attention to poetry.
As always, my words are set in boldface while ChatGPT's are plain.
The story
Give me a poem based on this anecdote:
And they told me one that concerns me directly, the one about Pope Francis in America. It goes something like this: As soon as he arrives at the airport in New York for his apostolic journey in the United States, Pope Francis finds an enormous limousine waiting for him. He is rather embarrassed by that magnificent splendor, but then thinks that it has been ages since he last drove, and never a vehicle of that kind, and he thinks to himself: OK, when will I get another chance? He looks at the limousine and says to the driver, “You couldn’t let me try it out, could you?” “Look, I’m really sorry, Your Holiness,” replies the driver, “but I really can’t, you know, there are rules and regulations.”
But you know what they say, how the pope is when he gets something into his head … in short, he insists and insists, until the driver gives in. So Pope Francis gets behind the steering wheel, on one of those enormous highways, and he begins to enjoy it, presses down on the accelerator, going 50 miles per hour, 80, 120 … until he hears a siren, and a police car pulls up beside him and stops him. A young policeman comes up to the darkened window. The pope rather nervously lowers it and the policeman turns white. “Excuse me a moment,” he says, and goes back to his vehicle to call headquarters. “Boss, I think I have a problem.”
“What problem?” asks the chief.
“Well, I’ve stopped a car for speeding, but there’s a guy in there who’s really important.” “How important? Is he the mayor?”
“No, no, boss … more than the mayor.”
“And more than the mayor, who is there? The governor?”
“No, no, more. …”
“But he can’t be the president?”
“More, I reckon. …”
“And who can be more important than the president?”
“Look, boss, I don’t know exactly who he is, all I can tell you is that it’s the pope who is driving him!”
Here’s a poem inspired by the anecdote about Pope Francis:
The Pontiff’s Joyride
Amidst the roar of jets, the skyline’s gleam,
The Pope disembarks, his cassock pristine.
A limousine awaits, a glistening beast,
The epitome of splendor, to say the least.But Francis, humble shepherd, eyes the prize—
A wheel to hold, the highway’s wide surprise.
“It’s been so long since I’ve driven,” he grins,
“Why not let me try, where’s the harm in sins?”The driver hesitates, protocol bound,
Yet Francis insists, his joy profound.
So off they go, the Pope at the helm,
Through asphalt seas, his private realm.He floors the gas, the engine roars,
Wind sweeping through limousine doors.
Fifty, eighty, one-twenty, he flies,
Till sirens wail beneath the darkening skies.A cop approaches, badge aglow,
But freezes at the window’s show.
"Boss, I’ve got someone bigger than the mayor.
No, not the governor. No, not the player."“More than the President?” the chief does implore.
“I don’t know who’s driving, but the Pope’s on board!”
The chief, stunned silent, says at last,
“Let him go—this ticket’s above my class!”So Francis smiles, a child at play,
His sacred joy in the thrill of the day.
Not just a shepherd of the flock,
But a driver who’s still breaking the clock.
* * * * *
Notice, however, that FTH has missed the point of the anecdote. If the Pope is merely the driver of someone’s limousine, who could that possibly be? Let’s set that aside for now and come back to it.
Subscribe to:
Posts (Atom)