Category Archives: Web Article Review

what’s new with learning curves?

At least since reading some early Singularity-adjacent publications by Vernor Vinge, Ray Kurzweil, and Bill Joy, I’ve been interested in learning curves. (And for the record, the topic and these authors were were not considered politically “right wing” or even political at all at the time.) Progress, at least in certain technologies, tends to be exponential over time. This clearly applies to computer technology, where there are short product cycles, the needed infrastructure is more or less in place and/or can adapt as the technology is scaled/commercialized, and legal and institutional barriers to change are relatively low. Technologies with “recipes”, like chemicals, drugs, seeds and other agricultural technologies, might be other examples. For these we have the patent system to actually try to slow down scaling and commercialization to the pace of innovation. With energy technology, learning curves seem to play out over much longer periods of time because while available technology changes rapidly, our system tends to be locked into long-lived infrastructure that can only change slowly. So new energy technology rolls out slowly as it is scaled up and commercialized over decades. There are also entities with enormous political and propaganda power that fight tooth and nail to keep us locked into obsolete technologies and infrastructure that fit into their historical (and profitable) business models. Now when you get to other technologies, like transportation and housing, public policy, legal and institutional barriers are dominant and tend to retard or even prevent progress. Rollout is so hard that while there are pockets of innovation, many don’t see the light of day or don’t spread from the local/pilot scale, even if they are successful at this scale. These also vary by location and jurisdiction, so that progress is very uneven geographically. These are my thoughts anyway. As for what’s new, here’s a journal article from Advances in Applied Energy.

Variability of technology learning rates

Climate and energy policy analysts and researchers often forecast the cost of low-carbon energy technologies using Wright’s model of technological innovation. The learning rate, i.e., the percentage cost reduction per doubling of cumulative production, is assumed constant in this model. Here, we analyze the relationship between cost and scale of production for 87 technologies in the Performance Curve Database spanning multiple sectors. We find that stepwise changes in learning rates provide a better fit for 58 of these technologies and produce forecasts with equal or significantly lower errors compared to constant learning rates for 36 and 30 technologies, respectively. While costs generally decrease with increasing production, past learning rates are not good predictors of future learning rates. We show that these results affect technological change projections in the short and long term, focusing on three key mitigation technologies: solar photovoltaics, wind power, and lithium-ion batteries. We suggest that investment in early-stage technologies nearing cost-competitiveness, combined with techno-economic analysis and decision-making under uncertainty methods, can help mitigate the impact of uncertainty in projections of future technology cost.

This blog in Construction Physics has a deeper dive across more industries, and discusses at least one large data set that is available for analysis. If you could accurately predict learning rates (and successful scaling/commercialization rates) for specific technologies based on known factors, then theoretically you could fine-tune policies and incentives to increase the rate of progress in the technologies you want. So this is an area of research that could boost all other areas of research and progress.

evidence for the return on (U.S.) government non-defense R&D

This 2024 report from the Dallas Fed provides very clear evidence of the positive returns from past U.S. government research and development funding.

Total factor productivity is a noisy but generally accepted measure of the amount of GDP/productivity growth that is due to innovation rather than increases in inputs. Summary: The return CAUSED BY non-defense R&D spending is 140-210% over 8-12 years, which is higher than investments in infrastructure (which still provide a positive return) and defense R&D (NO CAUSAL EFFECT IDENTIFIED).

Since it’s noisy, maybe I would smooth it in the graph above, but nonetheless there is a very clear relationship between falling R&D spending and falling economic growth. Conversely, if you wanted to intentionally reduce growth and innovation in our economy, a good way to do that would be to reduce R&D spending. Another implication is that if R&D spending on weapons and war does NOT provide as great benefits, there is an opportunity cost to spending your R&D money on weapons and war rather than peaceful or at least dual-use technologies. So it’s pretty clear the actions of the current US administration (drastically cutting R&D spending and shifting it from civilian to military applications) do not match their stated intentions to boost economic growth.

toilet rats, and a Singaporean perspective on Asia at the end of 2025

The two things in my title are only loosely related, and here’s how. The Guardian has a gleeful article about toilet rats in the (US) state of Washington. Indeed, this does seem like a pretty good indicator of US decline. Nonetheless, I have one personal experience with a toilet rat, and it was in Singapore. Older-style public restrooms (confusingly for American tourists, called “toilets” as they are throughout most of the world) sometimes have squat toilets flush with the floor rather than western toilets that you sit on. This is a traditional Asian style of toilet, only the modern squat toilet is collected to a modern sewer system rather than just a hole in the ground. Anyway, I was in the restroom/bathroom/toilet when a rat came out of a little hole in the floor, either not noticing or not caring that I was there. I stamped my foot just to let the rat know that I was by far the larger and dominant mammal in the room, and the rat reacted by diving directly into the toilet and down the drain. And it was gone. So I assume it just swam for a bit until it got somewhere with air, and returned to whatever it was doing after I left. Anyway, the advice in Washington State is if you see a rat in your toilet you are supposed to flush the toilet or close the toilet lid. I would not do either of these things because rats are FAST, they are afraid of people, and they have sharp teeth. I think I would calmly close the bathroom door and just peak in after an hour or two to see if it chose to go back where it came from. And I might keep the toilet lid closed after that.

I have been in actual sewers in the United States, and I have seen rats. Sewer rats are plentiful. They don’t want anything to do with us humans, they just scurry away if they see us coming. I have never seen or heard of one coming into a person’s house through the plumbing. So the article seems a bit alarmist to me. Mice are another story. I just wiped some suspected mouse poop off my kitchen counter this morning, which is gross and a disease risk. Anyway…

The connection between Singapore and rats is that Singapore, which has a reputation as possibly the world’s most sparkling clean very dense city, is not perfect. There is trash and there are rats, like any other city. Singapore, and its experts, have a certain self-endowed swagger. Now Singapore really is pretty clean, and its experts really are pretty smart, which brings me to my point that Singapore is pretty good but not perfect. So anyway, here is what a Singaporean expert, George Yeo, with a lot of credentials says is going on in Asia at the end of 2025. It is one person’s opinion but also a uniquely Asian perspective and different from what we hear from the US government/media/nonprofit “blob”. The article is from Hong Kong-based South China Morning Post by way of Yahoo, so also keep in mind the possibility of censorship and/or self-censorship (as you also certainly should with the US blob).

  • If it ever becomes clear that the US will not continue to support Taiwan, China will take over Taiwan. Taiwan will not fight a war it will obviously lose leading to its own destruction and loss of a generation of its young people. Right now the US and China are sort of avoiding talking about this and trying to actively suppress others (like politicians in Taiwan and Japan) from talking about it, and this is essentially a continuation of the long-term status quo.
  • When we hear about large weapons sales from the US to Taiwan, there is an unspoken arrangement that these weapons are considered defensive and do not cross a certain line. The US is seen to be supporting Taiwan, China is seen to be outraged, and both sides get their propaganda win without serious escalation. Again, it’s the long-term status quo.
  • Another reason weapons will not cross a certain line is that the US will not provide weapons where it has a technological lead. “Knowing that many Taiwanese are blue, the US cannot be sure that advanced technology supplied to Taiwan will not quickly leak into China. The military technology supplied to Taiwan is technology the US can afford to lose to China.” [Wikipedia: “The Pan-Blue Coalition, Pan-Blue force, or Pan-Blue group is a political coalition in the Republic of China (Taiwan) consisting of the Kuomintang (KMT), the People First Party (PFP), the New Party (CNP), the Non-Partisan Solidarity Union (NPSU), and the Young China Party (YCP). The name comes from the party color of the Kuomintang. Regarding the political status of Taiwan, the coalition primarily maintains that the Republic of China instead of the People’s Republic of China is the legitimate government of China. It also favors a Chinese and Taiwanese dual identity over an exclusive Taiwanese identity and backs greater friendly exchange with mainland China, as opposed to the Pan-Green Coalition which opposes Chinese identity in Taiwan.” This also, if I am not mistaken, is the status quo position going back many decades.]
  • “There is growing realisation that the road to independence is a dead end…If the young people of Taiwan build their hopes on an illusion – as the young people in Hong Kong once did – it will only lead to tragedy…Taiwan can enjoy more autonomy by negotiating now rather than waiting another 10 years.”
  • Taiwan has economic and industrial strengths that China would like to maintain and benefit from after a hypothetical reunification [which to me, would seem to discourage any full-out military onslaught on a major urban and industrial city]. This is similar to the situation with Hong Kong, where there is some degree of autonomy and the situation is short of full integration [but in my words – obviously much more limited in terms of political freedom than it was in the past].
  • “How can they ever forget that it was Japan’s aggression which separated Taiwan from the mainland in the first place?” [Interestingly, Taiwanese I know tend to have relatively warm feelings towards Japan. I had to refresh on the history again thanks to Wikipedia – Taiwan was occupied/colonized/ruled by Japan from 1895 to 1945. So it was not invaded and dominated in the 1930s and 40s like much of the rest of China and Southeast Asia, and people there were not mistreated as badly.]
  • On “rare earths” we may hear about the US and its allies developing other sources of rare earths, but there are certain “heavy rare earths” which only China produces, and which are critical to industries around the world. Yeo says China could have played this card at any time in the past, and did so only reluctantly as a bargaining chip in response to the recent round of trade disruptions initiated by the Orange Baboon-Ass God [my words, although he does go into a tangent about the Monkey King which is a Chinese/Buddhist legendary epic. Maybe this is actually a very subtle swipe at his dipshit highness.] Yeo sees this situation as a form of mutually assured destruction where it would be irrational for either side to escalate further.
  • The reality of the US government debt is that if there is a severe contraction or reduction in the growth rate, it would at some point have to print money to service its debt. Central banks around the world are buying gold and diversifying away from the US dollar because of this risk.
  • “He [His Orange-Ass Highness] recognises that the US cannot dominate the world the way it used to in the past. The US hasn’t got the financial power or the manufacturing capability. So it has to retreat some and consolidate around its own core and concentrate on healing itself.” The US knows it is overextended globally, and this is what the bluster over exerting itself in the Western Hemisphere is about. It won’t be able to continue exerting itself globally by sheer power and force, so it is retreating particularly from Asia while still trying to look tough. [Maybe, but aren’t there still the 800+ military bases around the world? And why would we antagonize allies if we are in a position of weakness? I am just saying this is irrational, but I admit that ideology can Trump rationality.]
  • He doesn’t see disputes between China and Vietnam or China and the Philippines in the South China Sea “boiling over”.

construction productivity

Construction Physics has a deep dive on construction productivity around the world. We hear about the overall slowdown in productivity growth worldwide since the 1970s or so, but in the construction industry the trend is essentially stagnation even compared to other industries. The U.S. is a historical leader in absolute productivity but has actually managed a productivity decline compared to modest growth in most other countries studied. That said, there are no countries where the growth is particularly spectacular. Developing countries have managed to grow productivity faster, but that is essentially catching up. It talks a lot about the challenges of measuring productivity, suggesting that just focusing on cost might be the better way to go.

This article doesn’t go deep into potential solutions. Prefabrication of components in factories is talked about a lot, because manufacturing productivity gains have been much more dramatic than construction, which on its face is manufacturing in a much less controlled environment. But prefabrication and modularity have been worked on for a long time and delivered only modest gains. More competition and less corruption in procurement are certainly good things, but these too seem to deliver only modest improvement. Many developed countries in Asia and the Middle East use labor from developing countries, and this seems to work for them but doesn’t deliver large gains I suppose because the lower-wage workers are less skilled and less productive. Streamlining permitting and regulation is always talked about, and tends to fit certain political agendas, but there don’t seem to be enormous gains there. So governments and project teams seem to just pursue an all-of-the-above salad approach and the result is incremental gains or no gains at all. I’ve probably said this multiple times, but I think AI should be very good at construction scheduling. Add in real time inspection and comparison to the original plans using cameras and drones, and it should be possible to really reduce down time and waste in construction. I think there might be substantial potential gains on the horizon here. If I were in government, I might focus R&D funding, targeted procurement, and regulatory/financial incentives on this particular aspect.

Another thought though, is that low construction productivity is not a reason not to do construction. Both housing and infrastructure construction have long-lasting economic and quality of life benefits that go beyond just the immediate economic activity they generate in the construction sector itself. So maybe we should just pony up what they cost now, keep plugging away to try to make the modest gains, and stop worry so much about this.

what’s new with fusion?

One thing that’s new, according to the New York Times, is massive investment in R&D by the government of China. Meanwhile in the U.S., it’s more about startups and public-private partnerships. In at least one anecdotal case mentioned in the article, top scientists from China who have been based at top U.S. universities for decades are choosing to go home. The article suggests that American firms are often the first to achieve breakthroughs, but the Chinese system is better at scaling and commercializing them.

It seems like there is an opportunity to cooperate for the greater good here, no? That is not the way the political winds are blowing at the moment, of course. At least in this case, if our countries aren’t actually cooperating, they are not competing to weaponize the technology first. This technology was weaponized more than half a century ago, of course, and the quest ever since has been to learn how to control it for peaceful purposes. Of course, the joke is always that commercial fusion is two decades away, no matter what decade we are currently in. This article declines to give a clear time table for widespread commercialization, but it talks about a “pilot plant in the 2030s and 2040s”, so yep the rolling two decade projection seems to be holding.

AI investment compared to railway boom

The blog Urbanomics has a comparison of the current AI investment concentration to the 19th century railroad investment boom in England and the United States. In this particular case, the blogger neglected to provide the original source, which he or she normally does. Financial Times and Economist are typical sources. Anyway, here are some stats mentioned:

  • Peak “railway mania” in the UK was around the 1840s, and railroad investment accounted for around half of all investment at that time.
  • Between about 1830 and 1870 in the UK, railroad investment accounted for about 20% of all investment.
  • In the US, episodic railroad investment booms occurred in the 1840s and 1870s. Railroad investment at these times was around 40% of all investment. This accounted for GDP growth of about 6-10%.
  • The brief clip actually doesn’t tell us how much of total US investment in 2025 is directed to AI. But it accounts for GDP growth of around 2%.

These are interesting numbers, but I don’t think comparing 19th century and 21st century US GDP growth is a very good comparison. That is essentially comparing a fast-growing developing country to a slow-growing advanced economy. If I had to pick one or the other to live in, I would probably go with the one that has safe drinking water, antibiotics, vaccinations, relatively painless dentistry, and air conditioning.

Prospera, Trump, Thiel and the Honduras drug trafficking pardon

The former president of Honduras, Juan Hernandez, who was convicted of drug trafficking and then pardoned by Trump, has ties to Peter Thiel and the “charter city” Prospera. I keep tabs on Prospera, as I was initially interested in the charter city idea after hearing a lecture by Paul Romer (a Nobel prize winning economist). The original concept, which was about economic opportunity, efficiency, and clarity of mission, has morphed into something far more corrupt and ugly.

This governing arrangement was organized in part by Hernández, who served two terms as president of Honduras from 2014 until January 2022. Prospera was permitted to launch the charter city after Hernández allowed for the creation of semiautonomous Zones for Employment and Economic Development, or ZEDEs. The ZEDEs were then overseen by a committee that included three of Hernández’s underlings and several American libertarian activists.

So there you have it. I am not saying this is proof that Prospera is the reason for the pardon. But what we see in the press is a lot of puzzlement that the U.S. government can be simultaneously fighting an anti-drug war on the Venezuelan government and supporting a pro-drug former Honduran government. This is logically inconsistent. On the other hand, simultaneously supporting right-wing big business interests in Venezuela (think oil contracts if not outright ownership for US corporations) and right-wing big business interests in Honduras is logically consistent (not to mention corrupt and cynical).

what’s new in the JFK files?

What’s new is evidence that James Angleton at the CIA was personally tracking Oswald, and (separately, really), extensive ties between Angleton and the nuclear proliferation project of Israeli intelligence. The historical backdrop at this time, also based on evidence, is that JFK was actively and vocally resisting said nuclear proliferation. None of which seems to be a smoking gun with fresh fingerprints, just another party with motive and opportunity.

Though Angleton insisted that the agency was inattentive to Oswald and unaware of the purpose of his activities leading up to Dallas, it has since been disclosed through unclassified JFK assassination records that Angleton personally maintained a classified 201 intelligence/surveillance file on Oswald for the four years preceding Kennedy’s assassination, strictly controlling which officials inside the CIA were permitted to see it through compartmentalization.

Angleton committed perjury before the House Select Committee on Assassinations, claiming he knew almost nothing about Lee Harvey Oswald before the shooting. In another, Angleton concealed the fact that Oswald had visited the Cuban embassy in Mexico City—a visit the CIA publicly claimed it only discovered after the assassination. As Jefferson Morley, author of The Ghost: The Secret Life of CIA Spymaster James Jesus Angleton, explained, the  counter-intelligence chief “preferred to wait out the Warren Commission rather than explain the CIA’s knowledge of and interest in Oswald’s visit to the Cuban consulate” in Mexico…

At the very moment a U.S. president was seeking to restrict Israel’s nuclear ambitions and limit the political power of its lobby in Washington, the CIA official in control of the Oswald file was secretly sharing intelligence channels, assassination communications, and off-the-books operatives with Israel—and lying to both Congress and potentially some of his own CIA colleagues about it. The government spent 60 years redacting those facts and Americans have a right to know why.

I might ask how many agents (come on, yes, Oswald was a CIA agent/informant/collaborator of some sort) Angleton had files on. Hundreds? Thousands? or just a few? Did having a file folder with your name on it really make you special?

what to do after those holiday meals

I don’t put too much stock in online nutrition and fitness advice, but here is what at least one article (Fashion Beans) suggests in the day or two after overindulgence.

  • Day 1: Start by drinking a whole bunch of water to start flushing salt from the system. Delay caffeine intake for an hour or two (ha, no chance I would ever do this, I have my priorities.) Basically just eat protein, vegetables, and a little bit of vegetable-based fat the rest of the day (hopefully there are some of these amongst the leftovers.) Take a 10-minute walk after every meal (probably never a bad idea). Exercise, but only lightly (as defined by the fitness bro who probably wrote this post.) Give alcohol a break. Get a solid sleep (they say 7.5 hours, sounds a bit overprecise to me).
  • Day 2: Lots of yogurt and fruit and more protein for breakfast.

And that’s it. Sounds totally fine for general health advice. We know what we are supposed to do right? Sleep, exercise, vegetables, fruits, nuts, seeds, whole grains, protein, approved fats and oils in moderation. 0-1 alcoholic beverages per day and avoid sugar and processed foods almost entirely. Do this for a couple days and you will probably feel decent whether you behaved badly on day 0 or not, I would think. Behave badly for several days in a row, as at least I tend to do over holiday breaks and on vacation, and you might start to feel pretty crappy. So here’s the best piece of advice I can give: Do as I say, not as I do!

what’s next for (incremental improvement of commercial) AI

We normals are hearing in the media that the large language model approach to AI has run its course, that further scaling it up is prohibitive in terms of energy, and that there is an AI-hype-driven financial bubble ready to pop any moment. According to at least one blogger though, the big breakthrough happening right now is having these models “reason” internally before they give an answer.

Two of those leading engineers are: Julian Schrittwieser who helped teach AlphaGo how to play Go at a level never witnessed in human history and is now a lead researcher at Anthropic. And Łukasz Kaiser, who whilst at Google Brain, co‑authored the paper that launched the architecture now driving every major released model on “Attention is all you need”…

Kaiser, for his part, corrects time horizons. The category of work that still belongs unquestioned to humans is shrinking. He states, with a deep belief, that these AI systems will be able to do any labor task currently performed on a computer within a timeframe of five years! …

The question is not whether machines will pass some imagined threshold in the future, but what it means that they have already crossed thresholds we still debate as hypothetical. A society reacts to what it believes is true, not to what is true. When the prevailing public understanding is delayed by years, institutions are, by definition, operating in a prior decade.

We can model technological progress as a series of sequential, overlaid S-curves that have to overlap in just such a way to produce continuous exponential growth. At least some insiders are still thinking in terms of keeping this S-curve going, in a competition between companies and countries. And when we see a new technology break through into widespread public, commercial use, it has already been going in the lab for awhile. That used to be measured in decades, now it is months if these optimist insider voices are to be believed.

https://onepercentrule.substack.com/p/is-ai-on-a-new-trajectory