Bill Gates weighs in on AI

Leave a reply

Normally I like to read and skim for main ideas, then distill it down in my mind to a paragraph or a few bullet points. If I let an AI do this for me, I skip the thinking part which is the whole reason I do this thing. But I simply don’t have time to even skim this article by Bill Gates, so I did ask Gemini to summarize it for me. I feel like he wrote this for an international NGO-type audience, and it is written in fairly general terms which a few politicians might try to turn in to specific action, but I can’t imagine this moving the US political system much. Anyway, he says governments should (at least I’m summarizing Gemini’s summary, so I have to use my brain a little bit):

  • Update tax and benefit policy before unemployment starts to rise.
  • Update worker retraining programs before unemployment starts to rise.
  • Strengthen cybersecurity.
  • Use AI to accelerate research and development specifically in health care, education, agriculture and climate change adaptation.
  • International cooperation blah blah blah.

The silver lining I can come up with is that in the United States we have existing programs and policies that address almost all of these (Social Security, Medicare, Trade Adjustment Assistance, the massive intelligence/surveillance/cybersecurity cluster, NSF, NIH, USDA/NOAA to name just a handful. We need a value added tax – but let’s think of a catchy name for the concept – which we don’t have.) They may be inadequate or outdated, but there is a foundation to build on. Internationally, I think the world needs to figure out if the UN can be salvaged or if we need something new (or some strengthening or cobbling together of other groups/frameworks that already exist) to perform the communication, coordination, and conflict resolution functions that the UN has provided (imperfectly) for the last 70 years or so and that it recently seems unable to provide.

The Soviet Union drilled a 40,000 foot deep hole

Leave a reply

This fascinating article describes the scientific enterprise. Among the things they discovered:

  • “For decades, geologists had interpreted a prominent seismic boundary deep beneath the continents as the transition from granite to basalt—the dense volcanic rock that forms most of the ocean floor. But Kola never hit basalt. The supposed boundary was nowhere to be found. Instead, the drill continued through more granitic and metamorphic rock, showing that the seismic boundary did not mark a change in rock type at all. It marked a change in how the same rocks behaved under immense pressure and temperature.”
  • “Nearly 22,000 feet beneath the surface, drill cores contained microscopic fossils from two dozen species of plankton that had drifted through an ancient sea more than 2 billion years ago. This showed that even after billions of years of burial deep within the continental crust, unmistakable evidence of some of Earth’s earliest life had survived.”
  • “Between 16,000 and 23,000 feet, the borehole encountered saline water circulating through fractures in crystalline rocks that geologists had long considered essentially dry. Rather than groundwater that had somehow seeped miles underground, the fluids appear to have originated within the crust itself. As ancient minerals were transformed by heat and pressure, they released chemically bound water, while reactions between water and iron-rich rocks generated hydrogen gas, showing that the supposedly inert continental crust was still chemically active billions of years after it had formed.”

So did they journey to the center of the Earth? No – to put it in perspective, the article says 40,000 feet is “higher than the highest commercial flights”. I know this is true having been on such flights, and yet when you are up there in a tin can you have no idea how high you are. Usually you can’t see anything, but I my recent flight I happened to glance down and saw a really fascinating landscape – possibly the Gobi desert? The formations must have been truly massive because you have no sense of scale. Anyway, digging 40,000 feet gets you “barely a third of the way through the continental crust beneath northern Europe.” What this tells me is that not only is the Earth massive, the atmosphere is a teeny, tiny film clinging to its surface.

What age does intelligence peak?

Leave a reply

“Fluid intelligence“, which is your ability to solve problems without applying outside information, peaks around age 20. I find this discouraging. But then “crystalized intelligence”, which is your ability to apply outside information to problems, does not decline until later in life (50, 60, 70? there are a lot of numbers in this article.) I find this encouraging as I am entering this window now. It also suggests to me that multi-generational teams may be good at problem solving if we can get past some of the cultural issues (seriously, what is with the avoidance of eye contact by 20-somethings today?)

The article goes on to describe measurement errors and biases that have affected past studies. A big one is that when you test the intelligence of a sample of people of different ages all at once, you get one answer, whereas if you follow the same people for decades and test them repeatedly using the same methods, you get a different answer. The latter approach is the better one, but of course it is difficult because it takes 50 years to carry out an experiment.

sea level rise projections

Leave a reply

A new data set aggregates sea level rise projections, of which there have been many.

Evolution of Global Sea-Level Rise Projections and Their Uncertainty

Recognizing this challenge, Garner et al. (2018)—hereafter, G18—developed a database of 21st century GMSL projections (A. J. Garner et al., 2018). An analysis of the G18 database generated several findings, including that the range of 21st century GMSL projections has varied greatly over time, despite central estimates being relatively consistent (A. J. Garner et al., 2018), and that upper estimates of GMSL rise from individual studies often exceeded upper bounds provided in Intergovernmental Panel on Climate Change (IPCC) assessments (A. J. Garner et al., 2018).

Eight years have elapsed since the publication of the G18 database and analysis, during which two additional IPCC reports incorporating new GMSL projections have been published (Fox-Kemper et al., 2021; Oppenheimer et al., 2019). To accommodate these new reports and other scientific developments and individual studies (e.g., Bamber et al., 2019; Brown et al., 2021; Grandey et al., 2024; Hamlington et al., 2024; Horton et al., 2020; Palmer et al., 2024; Weeks et al., 2025), we have updated and expanded the original G18 database of 21st century GMSL projections. The updated database (hereafter, G26 database) includes GMSL projections published through 31 December 2025 for a wide variety of time horizons—including decadal projections with end years throughout the 21st century, as well as longer term, multi-century projections. Other updates to the database include the addition of new fields, such as the country of origin for each GMSL projection study. These updates allow the assessment of the ways in which sea level science has evolved in recent years, but also to perform new analyses that offer additional insights into the history of the science, providing vital context for researchers and decision makers as they contend with challenges presented by the numerous studies and methodological choices in a deeply uncertain field.

exercise and dementia

Leave a reply

Exercise generally lowers dementia risk – hooray! But there are nuances – exercise performed in the context of work does not seem to lower dementia risk. This suggests to me either that stress plays a huge role, or that people with active jobs are actually doing more than the optimal amount of exercise. Maybe both. Another thing that surprised me here is that active commutes are not shown in the data to reduce dementia risk. That’s a shame, because I want people who shift from a car commute to an active walking or biking commute to be healthier. Maybe stress plays a big role here too – caused by real and perceived physical safety risks, noise pollution, and air pollution. If so, these are things we can try to mitigate through better design of our urban infrastructure.

the rise of electoral democracy

Leave a reply

Our World in Data shows is that, while there has indeed been a rollback in the number of democracies and number of people living in democracies over the last twenty years, the trend over the last 250 years or so have been a sharp and relentless uptick in electoral democracies. The US and UK were the first in the late 1700s, but New Zealand was the first to provide true universal suffrage.

I’ve been thinking, there are plenty of examples of rational and reasonably well functioning non-democracies in the world. And there are plenty of messy, imperfect, dysfunctional electoral democracies for sure. The article points out that a lot of electoral democracies exist on paper but have not had truly competitive elections in recent history. But we know the long-term relationship between economic growth and quality of life tracks closely with the rise of democracy, and we know that electoral democracy gives us some insurance against being trapped permanently in the truly terrible governance situations that sometimes arise in both democracies and non-democracies.

Waymo using Chinese EVs

Leave a reply

I’ve seen Waymo testing these vehicles around Philadelphia. It turns out these are Chinese-made, but Waymo adds the self-driving technology itself after importing them. Even after paying the tariff, they are about half the price of the Jaguar vehicles Waymo has been using.

https://electrek.co/2026/08/20/waymo-zeekr-ojai-chinese-ev-robotaxi-all-riders/

Polymarket has predictions of when Waymo will launch in various US cities, and these have ticked up significantly compared to the last time I looked. Chances of launching by end of 2026 (Sunday, August 23, 2026 as I write):

  • Operating now: San Francisco, Phoenix, Los Angeles, Miami, Dallas, Houston, San Antonio, Orlando, Austin, Atlanta (the last two dispatched by Uber)
  • > 75%: Tampa, San Diego, Denver, Las Vegas
  • 50-75%: Charlotte, Sacramento
  • 25-50%: Philadelphia, Detroit, Baltimore
  • < 25%: New York, London, Minneapolis, Seattle, DC, Portland, Boston

I’m glad to see Philadelphia with close to an even chance of launching by the end of the year. I personally wouldn’t bet on those odds, but they tell me that a year from now I will most likely to able to hail an automated taxi in my city, which will increase my safety and the safety of my children by roughly a factor of 10. And that doesn’t count reduced risk for pedestrians or the calming effect these vehicles are likely to have on human drivers.

August 2026 in Review

Leave a reply

Most frightening and/or depressing story: There are some dire projections of severe global famine in the next year. This would be due to the Super El Nino affecting important food-growing regions, closure of important grain-exporting ports in Ukraine, and the sudden loss/price spikes for fossil fuels and fertilizers that have already disrupted planting and growing seasons over the past six months, and will continue to do so. It will be similar to an 1877 El Nino that killed over 50 million people. I am a bit skeptical of the most extreme impact projections though because they do not seem to be a majority view, and the collective view of people participating in financial markets does not seem to be anticipating anything this bad. Still, it seems likely that regional famines and high food prices could affect a lot of people, and the worst projections would represent an existential threat that should not be ignored.

Most hopeful story: There is an effective malaria vaccine in use in Africa. It requires four doses, which hampers widespread implementation, but hopefully the numbers of doses will be reduced, eventually to one. In the United States, amid an embarrassing political dust-up over the measles vaccine (happening to involve my state’s governor, who does NOT at this point make the “top 5” list mentioned in the next paragraph), the FDA has quietly approved an mRNA-based flu vaccine.

Most interesting story, that was not particularly frightening or hopeful, or perhaps was a mixture of both: Election Betting Odds provides aggregated stats across the various prediction market platforms for a few categories including U.S. Congressional and Presidential elections. As I write (September 1, 2026), it is telling me there is a 51% chance the Democrats will take majority control of the U.S. Senate this fall (along with 84% chance of House control). The top 5 2028 Presidential contenders are Vance (21%), Occasio-Cortez (12%), Ossoff (11%), Rubio (10%), and Newsom (9%). I suspect at this point the latter is similar to a poll of “which of these names have you heard of?”. It might be a fairly good prediction of who will actually run, however.

the 1877 Super El Nino

Leave a reply

A Super El Nino in 1877 led to a famine that killed an estimated 50 million people, between 3 and 4 percent of the world’s population at the time. Deaths were concentrated in developing countries including China, India, and Brazil, which I am assuming would have had large populations of rural subsistence farmers at the time.

The article talks about how the impending 2026-2027 El Nino may be comparable in terms of heat and drought, but we have much better forecasting of what to expect and technology to deal with it. The article stops short of discussing how to actually deal with it, but I would assume some combination of stockpiling food in advance, international trade, international coordination and assistance, and (temporarily boosted, long-term unsustainable) irrigation. Also not discussed in the article are the self-inflicted geopolitical events (both the Middle East and Ukraine, which is an important grain exporter) that are making growing and trading food more challenging, right when this is most harmful. You can provide subsidies for farmers and consumers if prices spike, but at some point that will not help if there are really absolute shortages – if that happens, let’s hope it’s shortages of preferred foods rather than a shortage of absolute calories.

Yougov poll of support for bike lanes in US

Leave a reply

A yougov poll of about 1100 Americans shows that people support bike lanes in their communities by a pretty wide margin. Now, you could say this is a small sample size and people who ride bikes might be more likely to have answered. But, it is weighted for demographics and respondents were asked whether they own a car, own a bike, and how often they use these modes of transportation. There are plenty of respondents who do not own or ride bikes. And by a large margin, people support bike lanes. Not only that, they support separated, protected bike lanes, they are willing to give up some parking (but not car travel lanes) to add bike lanes, and they support (the only morally correct view in my opinion) that it should be legal to ride bikes on sidewalks when there are no bike lanes.

Respondents are grudgingly in favor of allowing e-bikes in bike lanes, but majorities oppose scooters, rollerbladers, skateboarders, robots, and people in wheelchairs. Sidewalks ideally should be designed and maintained to support some of these uses, but we all know that does not always happen when responsibility for sidewalks is often pinned on the individual property owner.

So my verdict is that, as is so often the case in our cities, the loudest, angriest voices opposing change create a disproportionate perception that people in general oppose change. And in the case of our streets and sidewalks, choices made by a few cynical public officials and reckless drivers can ruin it for all the rest of us.