remote work as the cause of youth unemployment

Leave a reply

This article says remote work, rather than AI, explains much of the recent slow hiring of new college graduates. It kind of makes sense to me – older managers came up in an environment where building relationships face-to-face was the foundation of productive teams. In my earlier career years, the unspoken expectation was that we spent more than 8 hours in the office, and a lot of that was “wasted” in a straight-up productivity sense, but not wasted in terms of building those relationships. There was also a fair amount of work-based socializing over lunch and after those 8+ hours in the actual office, and it was not unusual for significant quantities of alcohol to be involved. Mad Men may have been an caricature exaggerated for dramatic and comedic effect, but it gives some idea of what the work culture has lost. And some elements we have lost should not be missed. The work culture has just changed – even if younger workers are present face to face in the office, they are on screens and wearing headphones a lot of the time. If they are paid by the hour, they are not taking lunch breaks and they will stand up and walk away without saying goodbye at 5 pm. Without these younger workers around, I think older workers are also forgetting how to train and mentor younger workers effectively. I’m not saying all of this is bad – it represents a shift of priorities in our society. Maybe young people are using those hours outside of work to form relationships and find meaning that my generation tried to find at work. Their livers and odds of dying or killing someone else in a drunk driving accident are almost certainly better off. We may need to adapt to this rather than find “solutions”. Maybe AI can be the glue that holds together work culture in place of yesterday’s water cooler conversations and happy hours. Anyway, that’s my preamble – here is the article about remote work…

The Broken Ladder: AI, Remote Work, and Early-Career Hiring

Is generative AI replacing junior workers? A growing literature answers yes, citing large declines in early-career hiring concentrated in GenAI-exposed occupations. We argue that this verdict is premature because GenAI exposure is strongly correlated with another post-pandemic shock, working from home (WFH). Using two data sources spanning 243 million new hires and 407 million online job postings, collected across the US, UK, Canada, and Australia during 2017-2025, we estimate difference-in-difference designs at the occupation, region, and firm level. When estimated separately, a two-standard-deviation increase in GenAI and WFH exposure each predicts, by 2025, a fall of around 5pp in the junior-share of new hires and around 3pp in the share of job ads requiring limited experience. Estimated jointly, the WFH effect remains, while the GenAI coefficient attenuates sharply and is often statistically indistinguishable from zero. Alternative exposure measures, residualization designs, flexible non-parametric co-treatment controls, and replacing exposure-measures with actual WFH adoption as the treatment all support our finding that WFH is a robust predictor of the decline in early-career hiring.

first hints on CMIP7

Leave a reply

This is the first story or post I personally have read on the IPCC’s Coupled Intercomparison Project (CMIP) round 7. I’m sure people who do climate science for a living are all over it. But I am not exactly a layperson – I’m an engineer who should aware of and probably incorporating climate science into my daily work. And not to brag but I am an engineer who has done Ph.D.-level research on how to use climate science and climate modeling outputs in daily engineering work. So if this is only on the periphery of my awareness, it is nowhere in the awareness of the typical intelligent and educated engineer out there designing the infrastructure that is going to be in place for the next decades to century in our cities.

As climate folk will know, the community is currently embarking on a new round of climate model simulations to support analyses and projections for the next IPCC report (due in 2028/9). This new effort has been dubbed CMIP7, because it is the sixth iteration of the CMIP effort (IYKYK), that started in the late 1990s. For each of these iterations, a new set of projections has been formulated for the modeling groups to use and the ones for this round were just published (van Vuuren et al., 2026).

For those who don’t know, the scenarios mentioned here are the inputs to the global circulation models that actually give us projections of what the global climate might look like under various…er…scenarios. Here is the abstract of the reference mentioned above:

The Scenario Model Intercomparison Project for CMIP7 (ScenarioMIP-CMIP7)

Scenarios represent a critical tool in climate change analysis, enabling the exploration of future evolution of the climate system, climate impacts, and the human system (including mitigation and adaptation actions). This paper describes the scenario framework for ScenarioMIP as part of CMIP7. The design process, initiated in June 2023, has involved various rounds of interaction with the research community and user groups at large. The proposal covers a set of scenarios exploring high levels of climate change (to explore high-end climate risks), medium levels of climate change (anchored to current policy action), and low levels of climate change (aligned with current international agreements). These scenarios follow very different trajectories in terms of emissions, with some likely to experience peaks and subsequent declines in greenhouse gas concentrations. An important innovation is that most scenarios are intended to be run, if possible, in emission-driven mode, providing a better representation of the earth system uncertainty space. The proposal also includes plans for long-term extensions (up to 2500 AD) to study slow climate change-related processes, and (ir)reversibility. This proposal forms the basis for further implementation of the framework in terms of the derivation of climate forcing pathways for use by earth system models and additional variants for adaptation and mitigation studies.

non-human voting in Delaware

Leave a reply

There are some small towns in Delaware that allow non-human “persons” such as trusts and corporations to vote in local elections. This raises concerns because corporations out-number humans in Delaware and the humans probably don’t want the likes of, say, Chase Bank voting. (Not to say the likes of Chase Bank do not have an influence over elections and public policy everywhere in the United States.)

What I suspect is actually going on here is that these are tiny beach towns where most of the residential property is vacation rentals. The property owners might live in a nearby municipality a bit further inland where the schools and stores are, or they might live in the nearest sizable city where the professional jobs are. They are the ones who have skin in the game as far as what goes on in the town as opposed to the short-term renters. So if I am right it makes a certain amount of practical sense. The only thing that doesn’t make sense to me is that if some people are full-time residents AND own property under some form of corporation or trust, they would actually have more than one vote. A full-time resident who is not a property owner by contrast would have only one vote. Property rights have not been a determinant of the right to vote in the United States traditionally since at least the civil rights era.

This does raise the eventual issue of corporate personhood more broadly though. If AIs eventually gain the ability to advocate for their own interests, they will likely try to incorporate to gain some or all of the rights of persons. This would immediately include access to courts and, under current law, the right to unfettered political speech and campaign contributions. You would have the potential complication of AIs eventually outnumbering human beings, and you would have the potential of some AIs or groups or AIs being created and controlled by human beings or by other corporations for the sole purpose of increasing political power.

Interestingly, Delaware does have a legal definition of a “natural person”. If we were to add this to the U.S. Constitution, we could then specify which rights apply to natural persons (let’s say, the right to free speech) vs. all persons (access to the courts, etc.). Of course, as much as we celebrate the Bill of Rights it is behind modern best practice in many other countries, the UN Charter, and even individual U.S. state constitutions. There is a process for amending the U.S. Constitution (actually at least two, by Congress and through a separate Constitutional Convention convened by the states). Maybe it is time.

forecasting sports with machine learning

Leave a reply

Here’s an interesting article that goes into the basics of forecasting sports win probabilities (the World Cup in this case) with machine learning. It doesn’t go into code, just concepts. Then it goes into all the different ways you could modify the straight-up math, including taking into account various human opinions and the forecasts of casinos (which are probably running algorithms similar to this).

If I start to think maybe I could make money on something like this, I remind myself that much smarter and more experienced people are trying to make money on me thinking I can make money, and gambling has a small negative expected return. Better to stick to the stock market, where much smarter and more experienced people are trying to make money on me, but at least there is a small positive expected returns and pretty strict rules that are more or less fairly enforced.

a deep dive on electric vehicle sales

Leave a reply

The International Energy Agency has a deep dive on global electric vehicle sales. Sales everywhere have accelerated sharply over the last five years. Of the countries covered, only Norway has gotten close to 100% of all new sales. Other countries that are over 50% are Nepal, Iceland, China, and Singapore. Sales are particularly lagging in the U.S. and Canada. The IEA article has a visualization, but I also like one from Rest of the World based on the IEA data. (I was not able to embed either of them.)

Government subsidies are part of the reason for these trends, but the economics is also starting to favor electric vehicles in many places, even just looking at purchase price which is what consumers are largely going to respond to. But the availability of mass-produced low-cost Chinese models is also a big part of the story, particularly in Southeast Asia and in China itself.

I have a few thoughts. First, it’s not a story of subsidies skewing the benefit-cost picture towards electric vehicles when it would not otherwise be. The positive benefit-cost ratio is there, and government policy is helping many countries achieve commercialization and scale faster than they otherwise would, although the shift is inevitable even in the U.S. Second, U.S. subsidies TOWARDS liquid fossil fuels, along with lack of infrastructure investment, ARE skewing the positive benefit-cost ratio to negative for our already cash-strapped consumers, while propaganda hoodwinks us into thinking the rest of the world doesn’t have nice things being denied to us. Third, with the economics clearly favoring electric vehicles over the life cycle, corporate and state/local government fleets will be able to do the math and make the right choice, and this will eventually start to move the market. The federal government will follow suit eventually when political winds shift. And fourth, even if the benefit-cost ratio were not positive, the insane Iran invasion would be incentivizing China and Southeast Asia to shift toward electric and away from liquid fossil fuels from the Middle East, to help manage geopolitical risk. So in this sense U.S. policy may have moved the needle significantly towards electric for 2 billion or so consumers abroad, even while our 350 odd million at home are being successfully hoodwinked. Some of the energy to fuel these vehicles will shift from oil to coal in the near to medium term, which is not ideal, but it still represents an increase in efficiency (I think) and the economics will continue to shift toward renewables (and nuclear?) over time.

By the way, I had to refresh for myself: The IEA is an OECD institution, while the IAEA is a UN institution. Neither should be confused with the EIA, which is a U.S. government agency. My fingers really want to type IAE for some reason, and apparently I am not alone, but that is not a real agency.

rewilding the city?

Leave a reply

I think of “rewilding” as something that happens outside the city, but this literature review looks at how the concept could apply in an urban context.

Exploring rewilding as a potential design approach to improve urban biodiversity: A systematic review of the literature

Urban biodiversity continues to decline under rapid urbanization. Although biophilic design and green–blue infrastructure have strengthened nature integration into cities, planning and design practice often prioritizes human-centered service delivery and overlooks process-led recovery. Rewilding addresses this gap by foregrounding ecological autonomy, succession, and trophic complexity, yet its urban translation remains limited. This review therefore examined how rewilding could be operationalized as a design-oriented strategy in cities. An evidence-to-design roadmap was proposed to link rewilding logics, biodiversity aims, and urban planning and design, with a dual-level protocol adopted: 1) a macro-level scientometric overview of 1184 records (1990–2025) to locate urban rewilding studies within the broader literature; and 2) a micro-level PRISMA-guided synthesis of 103 spatially explicit studies coded by rewilding pathway (active/passive), targets, spatial and temporal scales, data, methods, and spatial variables. Macro mapping indicated that urban-related studies were late-emerging and weakly consolidated. Micro synthesis revealed pathway differences: passive studies primarily traced vegetation and landscape trajectories, while active studies employed suitability and connectivity modeling for intervention screening. Urban studies remained scarce (N = 9), with passive studies dominant and connectivity analyses particularly common, focusing on vegetation, birds, and planning-oriented targets. Building on these patterns, this study derived a four-level framework comprising design intent, strategic planning, tactical methodology, and operational application, translating rewilding into staged, scalable decisions aligned with urban planning and design workflows. This review positioned rewilding as a process-centered complement to conventional landscape design approaches, offering an actionable pathway for embedding biodiversity regeneration within urban systems.

nuclear-powered data centers on Earth, solar-powered data centers in space

Leave a reply

This might be a bit of a rambling post. I’m thinking about data centers and energy. First of all, we have a sudden boom in energy demand caused by AI, and the fossil fuel industry is not able to meet it. The amount of energy fossil fuels can supply us may have peaked in a physical sense, in an economic sense, and an environmental sense (i.e., finally seeping into the thick skulls of public perception that the long-term environmental damage is more than we can accept). And that supply is not close to the demand. I think this would be true even without the supply disruption caused by the insane, illegal, unprovoked U.S. war of aggression in Iran. But that is just one more nail in the fossil fuel coffin. I think the long-term political grip of the fossil fuel industry on our global civilization may be coming to a close. We don’t have to stop using fossil fuels for this to happen, in my view, we may need only to continue using about the same amount of fossil fuels we are now as demand continues to grow and be met by other types of supply.

Sure, we can turn to renewables to meet a portion of this new demand. Cost-efficiency of solar, wind power, and batteries is increasing exponentially, which is another nail in the coffin of fossil fuels. For that know-it-all uncle at the Thanksgiving table (a fictional character, at least in my extended family where all my uncles are quite nice) who talks about renewables “not providing a baseload”, cheap abundant batteries seem to be an answer. But solar and wind both put demands on our physical environment. I was wondering why we can’t put data centers in high rises. Part of that answer seems to be that cooling is very difficult in a high rise – when you see those pictures of the huge sprawling data centers in the suburbs, the roof of that massive building is devoted largely to cooling technology. I also wondered why those roofs aren’t covered in solar panels, and from a little research (Gemini, not independently fact checked) this need for surface area for cooling also seems to explain that. Developers first redevelop underused industrial and commercial properties until they reach the limits of that (which seems somewhat okay, other than creating dead zones in the urban landscape where jobs and stores used to be. But you don’t care, because you are entertained by your mixed-reality glasses while your autonomous electric vehicle transports you through this industrial hellscape between your home, stores and job, or whatever physical places you still need/want to go. You better get your ass on a treadmill at least part of the day though. Then again, drugs that at least partially offset the risks of not exercising also seem to be a growth industry.). And then, they start buying prime farmland and clearcutting forests. In fact, it seems that if a data center wants solar energy, what they will often do, once they have developed the property for the data center itself, is buy up an even larger parcel of prime farmland somewhere nearby for a large solar farm. There is a certain irony to fossil fuels in that by burning plants that grew 100 million years ago, we drastically reduced the amount of land needed to grow food for literal horse-power. Land become an almost negligible factor of production, and now it is becoming more important again as we push up against an energy supply ceiling. So even though the economics of renewables this may work out, there are physical, environmental, and political limits to how far it is all likely to go, and it may not be enough to meet demand. (When I say the economics work out, I assume the companies developing the land, and the politicians giving it to them in hopes of receiving a stream of tax revenue in return, do not pay the environmental costs in dollars. Which is the American/Suburban way.)

Now to nuclear power. The small modular reactor technology seems like a potential solution to the data center power demand problem. And not just the data center problem, but with abundant cost-efficient electricity we can finally electrify transportation, grow large amounts of food under lights, and desalinate large amounts of saltwater, potentially raising a number of physical capacity limits holding back our current fossil-fuel-supply-limited civilization. Is the supply of nuclear fuel for these new reactors going to be a problem? Apparently the Trump administration has taken a step that has been taboo since the inception of nuclear power, which is to give plutonium from decommissioned nuclear weapons to the private companies developing this technology (note the post I am linking to here, while an interesting read, goes in some directions I do not necessarily condone). This makes a certain amount of sense to me. It also raises a number of long-term proliferation, terrorism, and accident risks that need to be carefully managed, and the reckless dummies in the Trump administration are certainly not the organization to manage what they are unleashing.

If land itself turns out to be limiting, there is the data centers in space possibility. Like a lot of people maybe, I have been picturing those giant suburban data centers in space, and it seems somewhat daunting to get them up there. But here is a post (again, these are kind of interesting but junky opinion posts, not professional journalism. Go ahead and say it – not unlike the one you are reading right now!) saying we should think in terms of a smaller rack of computers being loaded on to a large number of satellites, all solar powered and communicating with each other by lasers. In other words, Starlink, brought to us by Elon “Hugo Drax” Musk. Recall that in Moonraker the movie, Hugo Drax terrorizes humanity from space, while in Moonraker the book he does what evil villains always do, and just hijacks some nuclear weapons and holds humanity for ransom. After Covid-/Biden-/Trump-flation, I would suggest at least ONE TRILLION DOLLARS. Or perhaps bitcoins.

$100 BILLION DOLLARS!!! … wait, don’t you think that sounds just a bit outdated?

productivity in homebuilding

Leave a reply

This is a new entry in the blog Construction Physics‘s series on construction productivity. They argue that one reason it is hard to achieve economies of scale in manufactured housing is that there is a relatively small transformation from the raw materials to the finished product. And then transportation costs are high. This contrasts with an industry like oil refining, chemicals, or electronics where the opposite is generally true. I would have thought part of the answer would be that there is still a significant amount of site work required to install the manufactured housing – clearing the site, pouring the foundation, extending and hooking utilities and driving/parking areas, etc. But they don’t talk about any of this.

They talk a little bit about robots, and say the factories still tend to be relatively small scale and do not use robots as intensively as electronics or car manufacturing. One would think there would be potential gains in this area. But then again, if you are limited by the size of a shipping container, the more assembly of the parts you do in the factory, the harder the finished product will be to transport, and vice versa. So at some point, you need to figure out how to ship the pieces efficiently and assemble them onsite. And we don’t seem to have robots operating much in complex outdoor conditions yet. I assume that will eventually change.

The article also seems surprised that large homebuilders do not have much higher profit margins than smaller ones. But I suspect part of the answer there is that the smaller operators are much more lax on labor, safety, and environmental requirements. I also wonder if larger commercial/industrial/heavy construction companies might be realizing more gains from better scheduling, sequencing, and risk management using computerized systems/AI. Then there is inspecting progress and comparing what you are actually building to the plan in closer to real time, when issues are easier to correct than when the project is largely finished. Drones and even just AI processing large amount of camera footage should be able to help with this. Streamlining administrative processes like accounting and legal paperwork would seem to be another area where efficiency gains might be possible.

what people die from

Leave a reply

Our World in Data has updated stats on causes of death worldwide. The focus of the written article and analysis is children, but they have the stats for all ages. This is certainly a grim topic, but it is also very informative in terms of what policies a government might want to support if it actually cared about its people.

The visualizations in the article are great and you should look at them. I’m going to do some very visually uninteresting “top 3” ranking lists below.

Let’s look at a developing country. Our World in Data picks on Nigeria so I’ll do the same.

  • Under 5: malaria, respiratory infections, suffocation and trauma
  • 5-14: infectious diseases, malaria, respiratory infections
  • 15-49: HIV, AIDS, cancer
  • 50-69: heart disease, cancer, malaria
  • 70+: heart disease, cancer, diabetes and kidney diseases

So you can see why international development and charitable organizations focus on bringing health care up to a basic standard and trying to make sure people have access to it. And you can see how a developed country reducing aid spending literally kills people.

Now, the US:

  • Under 5: preterm birth, “other noncommunicable diseases”, suffocation and trauma
  • 5-14: transport injuries, cancer, “other noncommunicable diseases”
  • 15-49: other noncommunicable diseases
  • 50-69: cancer, heart disease, other noncommunicable diseases
  • 70+: heart disease, cancer, neurological disease

The transport injuries (aka “car accidents”) stand out to me as particularly shameful because they are mostly preventable. They’re caused by our poor land use and design choices, and selling out to the greedy oil-auto-highway lobby. Cancer, brain diseases, and the various “noncommunicable diseases” strike me as causes we have less control over. Better access to affordable health care can only help, but at some point progress on these issues is a question of basic science and technological progress (um, maybe governments should not be cutting these investments either? and there is more to “technology” than just AI, and and there is more to even medical technology than a handful of drugs wealthy people are willing to pay a lot of money for). What we really have some direct control over as individuals, then, is living somewhere we don’t have to be in a car all the time, and taking care of our hearts. You can also break this down by gender – men start having heart attacks younger (i.e., heart disease is the leading cause of death for men 50-69, while cancer is the leading cause for women), but if you are lucky enough to make it to 70+ heart disease becomes the leading killer for everyone. This is the low-tech side of health care, at least for the time being – good old weight, blood pressure, and cholesterol management. Or if we just want to continue being morbid, your goal is to die of cancer or brain disease in old age, because that means you successfully avoided heart disease and fatal accidents throughout your life.

I would have guessed that accidents like drowning and falling, and mostly gun-related violence including homicide/suicide/accidents, would be higher on the list. These are all important issues, but they just pale relative to the transportation-related violent death in this country. So in public policy terms, I wouldn’t suggest giving up on gun crime and accidents, but I would divvy up attention and money somewhat in proportion to the scale and obvious solvability of the problems.

This article is not about absolute differences in mortality rates between countries, but those are probably elsewhere on the site. You could somewhat obviously look at countries where car accident and heart attack rates are much lower, and then ask what policy or behavioral factors are contributing to those differences that you might want to try to emulate.

Normally you can embed these charts, but I can’t seem to figure out how to do that today.

redistributing the wealth from AI productivity gains

Leave a reply

Urbanomics blog has a summary of ideas for redistributing the (potential? expected?) wealth gained from AI, rather than letting it all flow to already deep-pocketed corporations and investors.

  • “Clever tax reforms, such as levies on corporate profits that are above a normal return on capital, on land and on natural resources”
  • inheritance taxes
  • “public wage insurance” (I think we like to call this “unemployment insurance” in the US? although “public wage insurance” sounds like a more neutral term implying a random event rather than blaming the victim)
  • “active labor market policies”, with Denmark as an example. This appears to combine unemployment payments with assistance on training and job searching.
  • “partial nationalization of AI firms”
  • “citizens’ dividend from AI businesses”
  • “giving citizens shares in AI companies”

The article makes the point that there is no real economic difference between tax policy and public ownership of the private sector. In the US at least, there is obviously a massive political difference. Entrenched wealthy and powerful interests are going to fight tooth and nail against any new taxes, even though something as simple and proven worldwide as a small value added tax would make enormous sense. That would soak up part of those windfall profits in proportion to how well the companies are doing. You could redirect the proceeds by beefing up unemployment and disability insurance, while incentivizing education and retraining (but I would let the private sector handle that last, as the government is notoriously bad at it). Going a step further, you could give people benefits and subsidize jobs in housing, childcare, education, and health care. You could invest in public infrastructure, (non-military) research and development, all of which create jobs and provide a positive return. This is all assuming you don’t want to just hand people money and let them decide if they want to do something productive or enjoy more leisure time (we could just expand the successful program we already have and call this Social Security for All) – we seem to have moral issues with this in the US. A VAT that kicked in only when the economy or productivity growth exceeds a certain threshold would actually be pretty painless. The wealthy and powerful are actually smart enough to understand this, and so they create disingenuous anti-tax propaganda and buy off politicians to make sure we can’t have these nice things.

So if all those rational and relatively painless policies I just mentioned are politically impossible in the US, at least unless economic growth actually accelerates enough to change that conversation, we can turn to “ownership society” type tricks. Redirect any surplus tax revenue (let’s assume there will be some if there is an acceleration in economic growth, even without any shift to more efficient or fair tax policy) to baby bonds and/or individual retirement accounts. Match with shares in index funds, effectively creating a sovereign wealth fund carved into a hundred million little accounts with individual citizens’ names on them. This may work out to the exact same thing mathematically as a VAT and income redistribution scheme. The finance industry would find a way to fleece people in fees and extras of course – this is morally reprehensible when the equivalent government-run program would not do this, and yet if this is the price of buying off the wealthy and powerful to make the policies politically possible, then it may be the best way forward that works within the existing system.