robots and sidewalks

Leave a reply

We’re starting to see those Uber Eats robots on the sidewalks here, and they are spurring some renewed interest in the poor state of the city’s sidewalks. Now, I am surprised they don’t just design the things with bigger tires, as many parents have figured out that a “jogging stroller” can navigate our decrepit streets much better than a stroller with small tires designed for a fully ADA-compliant mall.

Anyway, since sidewalks here are the legal responsibility of private property owners, it occurs to me that Uber could start suing people if they wanted. It seems they would have a valid argument of economic damages. And trip-and-fall lawsuits, or the threat thereof, are a primary motivator of the sidewalk repairs that do happen, as this article in The Philadelphia Citizen points out.

But other cities have figured out better ways, which the article has a nice run-down of:

The 2022 ordinance enabled a straightforward deal: The City of Denver assumed all responsibility for sidewalks in exchange for property owners’ financial support. Denver began collecting a $150 annual fee from property owners and putting it towards citywide sidewalk maintenance. Last summer, Denver announced $75 million in repairs and construction from that fund.

Denver is a bold example, but it’s not alone. In many cities, residents don’t have to shoulder the full burden of sidewalk repairs. Chicago has a popular “shared cost sidewalk program” that has been running for decades and uses City resources to incentivize property owners to repair pedestrian walkways. Citizens pay a maximum of $2,400 — and typically around $1,000 — to participate, costs which are comparable to what a private contractor would charge. Cities like San Jose, NM and Memphis, TN have similar shared-cost programs.

And then there’s Washington, D.C., where a 6-year, $115 million effort to improve sidewalks is underway. The local government established a goal of building 8 miles of new sidewalks each year and repairing up to 40 miles of existing ones — work that aims to satisfy both the Vision Zero initiative and the Office of Disability Rights’ Olmstead Plan. In D.C., the land between a property and the curb is publicly owned, making it easier for the government to intervene. But the decision to prioritize sidewalks and enact a cohesive plan are actions that every city could seek to emulate.

Of course, people hate new taxes, and are only partially fooled by not-even-tax-deductible taxes disguised as service charges and fees. There’s also a complex issue this article doesn’t go into, which is that by making private property owners legally responsible for obviously public sidewalks, the city also makes them legally responsible for arguably public infrastructure underneath such as water, sewer, and gas pipes. But the point is that these programs solve a problem and give us value back in a more efficient manner than the non-system we have now.

Academic publishing beyond the current citation regime

Leave a reply

This blog post from someone named David Oks pines for a time before citations became a central pillar in the effort to advance scientific publishing. Thank you Mr. Oks for sharing your views, which got me thinking. I am not sure I agree 100%, because an important part of civilization and science is our ability to document the state of knowledge so that each successive generation can build on it. But the point is taken that the extreme focus on gaming this one metric (number of citations, citation scores of individual academics and journals) has become an end in itself, rather than a means for advancing civilization and science.

I do think AI can be very good at improving the state of the “literature review”. Every scientific article starts with a summary of literature on the topic, which the authors then typically build on (although, some articles are just literature review). A human author can spend years sifting through a vast amount of literature relevant to a topic, particularly a novel or interdisciplinary topic, trying to find those few needles in a haystack that are really the most, say, 100 highly relevant papers, and synthesize them into a foundation that can be built on it. I have done this, and it is actually a very fun thing to do (for my personality type, I suppose, not for everyone, but there are many like me…) Through this process, you gradually build and refine your own unique mental model on a subject which then can become the foundation for your personal unique contribution to human progress. However, even spending years, you can’t come close to looking at everything possibly relevant, and you can miss some of those needles. An AI should be able to look at literally the entire haystack, find and synthesize the needles, in a matter of minutes or at most hours, compared to months or years for a person. The only sad thing here is that the best mental models in the brains of the smartest humans might not be built in the same way. Overall though, I suspect the rate of progress can be increased significantly.

What is an alternative to the citation regime according to Mr. Oks?

So I suspect that we’ll have to fundamentally rethink the institutions of scientific life for the age of strong AI. Perhaps, as AI makes it possible to do much more science much more quickly, the culture of science will become more like the culture of engineering—faster, more collaborative, less interested in priority claims. In such a world, the most efficient unit of scientific contribution might be a living document, perhaps even just a GitHub repo: something with data, code, analysis, and a thin narrative layer that AI scientists could read, regenerate, or update as needed. And citations, in this world, could ultimately become obsolete. Journal articles would survive, though perhaps they’d become something closer to definitive pronouncements on major breakthroughs or on the state of knowledge in a given domain—a bit like what scientific books were before the rise of journals. In a world where science is much more productive than it is today, legitimacy will be the scarce factor in the production of useful scientific knowledge.

I am not sure I have experienced the culture of engineering described here in my own engineering career, but this may be referring to something more like “product development” rather than the public infrastructure and environmental planning type work I do. Anyway, my idea of AI-backed knowledge synthesis plus this idea of open data and code may be on to something.

e-scooter hate

Leave a reply

Here’s a Canadian e-scooter hit-piece. You definitely won’t find me on an e-scooter in Philadelphia, because both the outdated infrastructure and driver behavior make it much more risky than I could ever accept. But that suggests the solution: rethinking and modernizing infrastructure. E-scooters belong in the “bike lanes”, and bike lanes need to be modernized to accommodate all sorts of light personal vehicles. They need to be separated and physically protected from high-momentum vehicles designed for inter-city highway travel, and if they are combined with infrastructure for these high-momentum vehicles at all, they need their own signals at intersections so that the two are never, ever in conflict.

You need some enforcement, but we should remember that the core reason for all the traffic rules, licensing, registration, insurance, sobriety checkpoints, search and seizure, personal injury litigation, etc. is that cars are just so deadly. Yes, there are anecdotes of bikes and scooters running into each other and running into pedestrians and causing the occasional injury, but most of those are not going to be serious. So there is a question of how much we want to constrain personal freedom (as we do very significantly and appropriately with private cars) when it comes to much less dangerous forms of transportation (now, about those aggressive jogging strollers…)

I might allow autonomous golf carts traveling at low speeds (5-15 mph maybe) in the bike lanes, although today’s bike lanes are not wide enough for that. I might just ban private high-momentum human-driven vehicles from most streets and let the golf carts use them. Cities need this if forms other than cars are going to be viable forms of transportation in all kinds of weather for all kinds of people, and not just a from of recreation for the young and healthy. Of course, at this point the cyclists and e-scootists might decide they want to use the street too, and that might end up being okay. And then, especially if we don’t need so much space for parking as shared autonomous vehicles and human-powered transportation modes become more practical, we can turn part of what used to be the bike lane and sidewalk into gardens, affordable housing, or whatever a city needs. It’s a nice vision, and anything in the urban U.S. public infrastructure context has just taken much, much longer to progress than I would have thought in my youth. But I hold out some hope the breakthrough of autonomous electric vehicles, which is inevitable, could catalyze some faster change.

instructions for the AI scientist

Leave a reply

Here’s an example of how AI science could work. If you ask machine learning to review a data set from a system and predict the future behavior of the system, it can often do a good job, but the mathematical approach it is using to do that is mostly opaque. It is also divorced from any sense of the system structure, its interaction with the physical universe, and the physical laws governing it. But you can ask the computer to constrain itself within those known physical laws, and then it may be able to provide you insights on the structure and physical processes inside the system. So at that point, you and/or the computer should be able to form scientific hypotheses and test them against the data. This example from Water Resources Research is about soil moisture.

Coupling Intuitive Physics Into Deep Learning for Soil Moisture Flow Processes Learning

Soil water flow processes in the unsaturated zone support ecosystems and regulate water, energy, and biogeochemical cycles. Recently, deep learning (DL) approaches have significantly advanced soil moisture (SM) prediction tasks yet still challenging to interpret. It’s difficult to peer into the internal reasoning procedures of algorithms, let alone associate them with specific physical processes. Thus, DL alone is unlikely to satisfy soil hydrological modeling needs and cannot advance process understanding. Here, we present DPL-S (deep process learning for SM dynamics) approach, which couples intuitive physics into deep learning architecture as structural guidance to facilitate comprehensive surrogate modeling of soil water flow. DPL-S discretizes the SM state evolution into multiple sub-process effects (e.g., gravity, matric potential) at the intuitive physics level and abstracts them into format-specific and learnable tensors. By cascading state-action matrices in a differentiable end-to-end framework and enforcing penalties for physical inconsistencies, DPL-S enables a profound understanding of physical functions and scenes of soil water flow. Comprehensive numerical experiments including layered soil conditions and tests with in situ observations, demonstrate that it achieves reliable SM profile reconstruction with predictive performance comparable to the state-of-the-art DL model on supervised items. The internal inference of DPL-S is fully transparent and the tensor representations achieve strict physical realism under limited water content supervision, thus enabling continuous predictions like physical models during the testing period. The model’s flexibility, generalization, noise resistance, and large-sample diverse data synergies are also evaluated. This work represents a solid step toward learning hydrophysical processes from large data sets.

yes, the CIA is still meddling in other countries’ affairs (and also, because we appear to need it, a reminder that the word “peace” means NON-violence)

Leave a reply

If we need some confirmation that the CIA played a role in the January 2026 Iran street protests, here is Fox News journalist Trey Yingst seemingly quoting an interview with Donald Trump. The quote is ‘ “We sent guns to the protesters, a lot of them,” President Trump told me. “And I think the Kurds took the guns.” ‘

Now, this is an X video of a Fox News segment where an informal conversation with Trump is quoted. It’s Fox News, which of course is known to spin. But (1) it is a major news outlet that doesn’t usually lie outright, even if it spins and (2) this seems to be a respected professional journalist, not an opinion piece. So I give it some weight as having a significant probability of truth.

The CIA messing in other countries’ elections and opposition movements is not a new thing, of course, and it is not only the US that does this. European countries, Russia, and China certainly do it. In fact, the US did it in Iran in the 1950s, and that event is seen as a significant reason Iran and the Iran-US relationship are where they are today.

Maybe the invasion was intended to back up the protests, as Trump blustered at the time that the U.S. military was “locked and loaded”. So it makes me wonder if the protests broke out earlier than they were supposed to, when the US military was not ready, or if they broke out when they were supposed to but the US military was just not ready, or Trump just failed to pull the trigger at the planned moment. Nothing I am saying here justifies the illegal, unprovoked war of aggression on the sovereign nation of Iran. I am just saying it appears to be an illegal, unprovoked war of aggression that was also 100% incompetently handled. We are ruled by evil fools, not evil geniuses.

This also causes me to give more weight to the Russian claim that the CIA meddled in Ukraine’s affairs in 2014, stirring up an opposition movement that deposed a possibly fairly elected pro-Russian government. The back story on this is that first, a pro-Russian government was forced out by the street protests. But then there was an election, which brought a pro-Europe/US/NATO government to power, and was certified as free and fair by impartial international bodies. So far so good, but the pro-Russian parts of Ukraine largely did not participate because they were either under Russian occupation – Crimea – or occupied by pro-Russian local militia types – Donbas. So there was meddling on all sides, and much more direct and openly violent meddling by the Russian side. Then later, these events were used to justify the Russian invasion of the sovereign nation of Ukraine, which can have no legal or moral justification.

So if cooler heads ever prevail, we need to re-establish the idea of respect for soveignty. And the US could even go so far as to say it is not going to meddle in the affairs of other countries any more, other than through open diplomatic means and through international bodies. Doing this unilaterally might seem naive, since other countries would almost certainly continue their meddling – the classic prisoner’s dilemma, which also derails so many attempts at rational arms control. But when you consider that the meddling seems to lead to undesirable outcomes more often than not, maybe it would not be naive after all. We can cite any number of conflicts from the overthrow of the elected Iranian government in the 1950s, the mostly forgotten Indonesian genocide also in the 1950s which killed half a million people, support for the Taliban in the 1980s which led to 9/11, and name pretty much any country in Latin America. So my modest proposal is we just stop. Recommit to peace (but now thanks to the fool in the White House we have to actually state that this means NON-violence) and support for democracy and human rights through diplomacy and participation in legitimate international bodies.

Well, that turned into a rant I didn’t necessarily see coming. If you got this far, whether you agree or disagree, thanks for hearing me out!

How problematic is U.S. national debt?

Leave a reply

Here’s a plot from Gemini, not fact checked by me or any other human. Thanks Gemini!

I’ve always thought reporting “debt as % of GDP” is dumb. What really matters is how much interest payments on the debt are relative to the size of our economy. Or, in a more rational, less political environment, that is all that would matter – but in our real world politics matters a lot, and because politics limits our government’s ability to use taxes to pay the debt, debt payments as % of tax revenue also matter.

So…after World War II interest payments on the debt were very high, but this wasn’t a big deal because the economy was growing very quickly. In the 1980s and 1990s, interest payments spiked as interest rates spiked and growth slowed down. Eventually interest rates came down and got us out of that particular pickle. But now, from the plot we can see that current interest payments as a % of GDP are spiking to a similar level to how they did in the 1980s and 1990s. Interest rates are higher than they have been in recent decades, but not crazy high like in the 1980s. The difference really is the size of the debt relative to the economy. We can hope for faster growth to get us out of this one – there is some hope for AI-led productivity gains, but at the same time we have our government shooting itself in the foot by gutting research, development, and education spending, the historical underpinnings of our nation’s growth, while also blowing enormous sums on reckless, illegal wars of aggression with no end in site, and actually reducing taxes on affluent tax payers and corporations. We have inflation and interest rates both seemingly ramping up. So the situation does indeed seem pretty dire. Do I really even have to suggest solutions here? Sure, don’t stand in the way of the AI thing, but also don’t put all our eggs in that basket and do the opposite of all the obviously stoopid policies I just mentioned.

coal and China

Leave a reply

Yale Environment 360 explains the situation with coal use in China.

Paradoxically, China is at the same time the biggest installer of renewable energy, the biggest emitter of greenhouse gases, and the biggest user of coal. One explanation for this conundrum is a national concern over energy security: Coal is the only fossil fuel that China is not obliged to import, either through vulnerable pipelines or along sea routes that pass through precarious choke points like the Straits of Hormuz. China has an abundant supply of coal, boasting about 13.3 percent of the world’s recoverable coal reserves, and, importantly, it is the one fossil fuel that Chinese planners know will remain abundantly available, regardless of any tensions in China’s East Asia region or military action in the Middle East, the region that supplies China with nearly half its oil. This means that despite China’s role as a renewable energy superpower, coal has continued to play a leading role in its energy system. 

They talk about a decrease in “energy intensity”, which is energy use per unit of GDP. So the economy is growing, energy demand is growing, and renewables and battery technology are able to keep up with some but not all of that growth. Hydroelectric is a big part of their energy strategy, and that has been affected by drought recently. There are also complicated reasons why their grid is not run as efficiently as it could be.

My main impression is that it all sounds so…rational. Compared to the U.S. government which at the moment appears to be corrupt, immoral, and just bat-shit crazy.

Thermoeconomics in a Time of Monsters

Leave a reply

I am always trying to puzzle out how the financial and real economies are related, and how the loss of functioning ecosystems may eventually put the brakes on both – hopefully, in theory, with the financial markets acting as some sort of cushion where prices rise and allow some degree of adjustment that reduces the onset of actual physical shortages and severe hardship. Anyway, I heard an interview with Warwick Powell recently that seemed to cover some of this. Below is the Goodreads description of his most recent book (which I haven’t read).

Thermoeconomics in a Time of Monsters: Rethinking Theory, China and International Geopolitical Economy

In an era when the old world is dying and the new one struggles to be born, Thermoeconomics in a Time of Monsters delivers a groundbreaking reframing of political economy through the unforgiving laws of thermodynamics.

Warwick Powell argues that all social-economic systems are fundamentally energetic and metabolic. Their reproduction, stability and resilience are perpetually threatened by entropy—the natural tendency toward disorder and energetic dispersion. Successful societies survive and thrive through deliberate negentropic energetic renewal, efficiency gains, and institutional arrangements that convert energy surpluses into sustained order.

Building on classical concerns with value, surplus, production and circulation, Powell develops Systemic Exchange Value (SEV) theory. Money, information, supply chains and fictitious capital are reinterpreted as energetic phenomena. The result is a coherent ontology that integrates thermodynamics, endogenous money and information theory.

I often hear the claim that classical economics “ignores” or “doesn’t know about” land, energy, natural resources, or ecosystem services more generally as an input to production. I don’t think this is exactly true, they just choose to assume these factors are negligible compared to labor and capital, and they have some sound evidence that this assumption has been reasonable over the past couple centuries. What could change their minds is if the assumption becomes unreasonable because high prices or outright shortages of food, water, energy, and/or the physical environment’s ability to absorb our wastes becomes a large factor relative to labor and capital. Then maybe that will force a merger between the abstract models of economics and the iron laws of our actual physical universe. But is it possible that incorporating these factors in advance could help us head off the crisis? You might want to try to invent a parachute before you jump off the cliff, not while you’re in freefall.

https://www.goodreads.com/book/show/249361440-thermoeconomics-in-a-time-of-monsters

medieval castles and scalders

Leave a reply

Here’s a fun video on the defenses of medieval castles. Basically, they were designed to be very difficult to take militarily.

Sadly, “medieval scalder” is a job that has been made obsolete by modern life, replaced by something like “ballistic missile launch technician”, which itself is soon to be replaced by Terminators. Chris Farley and Macauley Culkin, 1991. RIP Chris Farley. Macauley Culkin is 45.

the future of scientific and engineering modeling

Leave a reply

I’ve been using AI to assist me with coding (R, in my case) since shortly after ChatGPT came out. In engineering, we tend to run off-the-shelf models that were of course written in some kind of code. Sometimes these are open access but often they are proprietary. The brutality of market discipline pretty much requires specialized off-the-shelf solutions in industry because customers are not going to be willing to pay for custom coding. The proprietary ones are even often preferred for legal/liability reasons. Anyway, the future of modeling appears to be humans providing a detailed specification to an AI agent which then follows it to do the coding, debug the coding, run the model, process and present the results. The humans have to be able to detect whether the results are BS, of course, at this point in history. One can imagine using a different agent or a more specialized agent to assist the humans in the bullshit-detection stage, that agent getting more independent over time, and so in a cycle. I wonder if we will be using agents to set up, run, and post-process the specialized models, or if things will trend toward just letting the agents write more fundamental code over time. Or maybe the specification will be what future scientists, engineers, and business people focus their efforts on, with translating that into 0s and 1s being basically a commodity done on the fly by AIs. This makes sense to me – the most crystal clear function of AI so far, in my view, is making it easier and easier for humans to communicate with computers in more abstract language, logic, and mathematical symbols.

Anyway, this example used something called Roo Code which included a couple versions of Claude Code along with some other agents, to run a fishery-related model. There is a peer reviewed article, but I also like this blog post and this example of a specification given to the agents.