According to one estimate, malaria deaths per year dropped from about 840,000 to about 440,000 between 2000 and 2015. The World Health Organization estimates that as of 2018 that number was about 405,000. The disease has been wiped out in many countries, and the methods for doing it are not complicated.

One of those methods is to sleep under insecticide-treated mosquito netting if you live in an area where you are in danger. The Against Malaria Foundation provides these mosquito nets, and has earned very high marks from the sorts of people who study these sorts of things.

Originally written in 2020

Although solar energy makes up a rapidly increasing fraction of global electricity, it is only immediately available when the sun is out. This creates various problems associated with the term duck curve. Those problems would go away if batteries were to get good enough: you store up energy when the sun is up, then use it when it's down.

And progress in batteries also will help determine the future of cars. Right now, battery power is more expensive than gasoline power, though battery power is getting cheaper fast: an 80% drop in prices since 2008, according to a New York times article from 2020-9-20. If the prices drop far enough, a battery-powered car will be as affordable as a combustion-driven vehicle. If they drop further, it's hard to imagine electric cars wouldn't dominate the market. The article quotes various experts who think battery power might reach financial parity with gasoline sometime around 2023 or 2025. It's worth checking in from time to time to see whether batteries continue to advance so rapidly.

Update (10 February 2023)

The NYT has published an article entitled Electric Vehicles Could Match Gasoline Cars on Price This Year, with the sub-heading Competition, government incentives and falling material prices are making battery-powered cars more affordable sooner than expected.

As long as we're conscious, we're breathing, with exceptions of four minutes or less. The breath is always there, and paying attention to it can help clear your head and calm you down, should you feel the need to be clearer-headed or calmer.

2020 - 2021-5-23

DuckDuckGo is a search engine that places anemphasis on user privacy, in an attempt to provide an alternative to Google's business model of extracting data about people who search.

DuckDuckGo continues to see rapid growth in its share of search space. As of May 2021, their website [link dead as of 2023] reports that about 97 million queries are being processed daily, which would suggest they're heading for 35 billion or more queries. Last year they fielded 24 billion, the year before 15, and before that 9, 6, 4, 3, 2, and 1 in 2013. In 2012 they had about 500 million; in 2011, 100 million; and in 2010, 16 million. Each year is seeing significantly more traffic than the one before. Over the last eight years, growth has been surprisingly close to an exponential 50%, year over year.

Still, Google is processing something like 4 billion searches per day -- estimates vary -- meaning that DuckDuckGo still handles a mere 2.5% or so of what Google does. Still a growth rate of 50% annual, carried forward, would see DuckDuckGo overtake Google around 2030.

However, there are some new developments that could conceivably shake things up further. Right now (2021-5-23), Google is in the crosshairs of some major antitrust legal action, continuing its long march through the court system which could conceivably undermine its business model of paying technology firms to push Google as a default setting to customers. If this is successful, DuckDuckGo might have an easier time. Alternately, DuckDuckGo might just run out of steam and stagnate or disappear. But it's worth watching to find out what happens next, especially for people who are concerned about what happens to private information online.

Update (Feb 2023)

I see a report from 20 June 2022 that duckduckgo traffic had declined from about 105 million queries per day in January 2022 to about 94 million in June 2022. That report linked out to duckduckgo.com/traffic, which appears to be no longer functioning on the duckduckgo website. Archive.org confirms that duckduckgo.com was reporting about 93 million daily searches as of June 20th. By July 2nd, the page had been altered to now longer show a running average of searched, but simply to show annual traffic in a manner which did not show the decline clearly. On July 5th, the decline was once again clearly displayed, with a for-July average of 88 million per day showing. Stats continued to be visible till sometime on or slightly before December 6th, at which point the stats URL began redirecting to the duckduckgo homepage. This does not seem like a positive development.

2020 or earlier

I used to think I understood the economy. Back around 2009, when the Fed started pushing the monetary base upward at a rapid pace, I knew that this would cause a dramatic inflation, and that the vast majority of economists were ignoring the simple facts about the money supply. By 2015, the monetary base had quadrupled (graph here). No spike in inflation. It turned out that mainstream economists understood the economy much better than I did as an eighteen-year-old. Since then, I've tried to pay much more careful attention to expert consensus.

Update (10 February 2023)

During Covid, we spent even more money than in the 2008-2009 crisis, and we did develop some inflation. It is reassuring to know that I'm not completely lost on where inflation comes from, although I can't pretend to know that much about it.

For just about any field of human study, there are lots of people out there who have made it their life's work to study it. You can save a lot of time and aggravation by listening to these people.

Effective altruism is a movement associated names like William McAskill, Toby Ord, and Peter Singer. These folks think it is important to try to quantify the amount of good that can be done with a given number of dollars in charitable work. Even if it's not possible to reduce doing good to simple equations, the effort to quantify the issues seems to me worth paying attention to. They're working on a variety of other interesting questions as well.

Interesting books about effective altruism or adjacent topics include William MacAskill's Doing Good Better; Peter Singer's The Most Good You Can Do, Famine, Affluence, and Morality, and The Life You Can Save; Toby Ord's The Precipice; Larissa MacFarquhar's Strangers Drowning; and Derek Parfit's Reasons and Persons.

2020 - 10 February 2023

Though we still await the Year of the Linux Desktop, open-source software has in some ways taken over the world, playing a major role in everything but the desktop. And desktop Linux has gotten pretty good. I originally typed this using the default text editor of Ubuntu to input Markdown into a text file. The editor highlighted the syntax automatically; I didn't have to do anything to set it up that way. It just knows I'm in Markdown and helps me out. When I want to use my Markdown files to build a website, Jekyll is there to do it for me. When I wanted to more smoothly manage my Markdown files, I switched to Obsidian, and then a series of open-source tooks replaced Jekyll for me.

Firefox, the default browser in Ubuntu, seems to do everything I want it to. While repeatedly checking election projections at fivethirtyeight.com in the lead-up to the 2020 elections, their feature tracking odds of winning for both candidates didn't seem to work on my phone's Chrome browser, but Firefox opened it fine. Even Chrome runs on an open-source core -- Chromium -- and that in turn is running on my phone's Android operating system, which is built on a modified Linux kernel.

Free and open-source software is just one genre of things that happen outside the realm of restrictive copyright. This larger set includes Wikipedia, which works in practice even if not in theory, and the massive and growing collection of books written in 1927 or earlier that one can find on the web. More recent books are more hit-or-miss, but are increasingly available to anyone who has googled "Sci-Hub" or "Library Genesis". Meanwhile, in a zone somewhere between the legality of old public domain books and the illegality of Library Genesis, the Internet Archive is experimenting with something they call "Controlled Digital Lending", which last time I checked was in legal limbo, awaiting court decisions.

This whole bundle of ways that people are building things outside of traditional copyright is worth paying attention to.

It is hard to find a more interesting collection of essays online than at gwern.net. Not only is it interesting, but it's got a great layout. And thanks to Gwern's decision to release the whole website under the CC0 license, anyone who wants to can reuse or adapt the layout freely.

How long things last is worth paying attention to. In a software context, everything is fighting against software rot. If you want something to last, you've got to account for software rot. Over at gwern.net, Gwern has some interesting thoughts about this. It is from reading Gwern's thoughts that I decided to build this site around a set of markdown files which are converted to relatively simple static HTML pages. I haven't -- so far -- followed Gwern into the world of Haskell, though.

More generally, the average thing is probably about halfway through its lifespan, almost by definition, which means the average thing will last about as long into the future as it has lasted so far. This is an extremely rough and unreliable rule of thumb, but something that's survived a thousand years has a decent chance of making it another thousand, while something that first happened yesterday might not stick around as long. Thinking too hard about this can lead to some weird places, like the Doomsday Argument.

More a booklet than a book, Sam Harris's Lying is an excellent summary of the case for why lying is bad. He takes a stronger anti-lying stance than most people do, and makes his case ably.

Copernicus may have been a vindicated contrarian, but for every Copernicus there is a Dick Johnson who thinks the earth is hollow and tries to drum up Congressional support for an expedition to the center. Dick Johnson was an actual Vice President of the United States, and he actually did try to drum up Congressional support to explore the hollow earth. Cranks outnumber geniuses, and so it's always a good idea to at least take a serious look at the majority opinion on any given topic. It's also a good idea to see whether the majority opinion among experts in the field differs from the majority opinion of non-experts.

Two hundred years ago, major sources of suffering for a lot of people included dying children, hunger, and contagious disease. Over time, those sorts of human suffering have greatly decreased in a lot of the world. As we increasingly manage to pick the low-hanging fruit of solving the basic material problems of life for more and more people, this leaves mental health issues taking up a larger and larger share of human suffering, even if average mental health is getting no worse. Tackling the most common mental health issues out there, especially depression and anxiety, is going to be one of the major challenges humanity faces going forward, if for no other reason than our increasing success at knocking out the other major challenges.

If you've read an interesting study but haven't looked to see whether there's a relevant meta-analysis, consider looking. It is particularly bad that most "science journalism" consists of summarizing single studies without giving the reader any broader context about other studies of the same thing or, especially, any meta-analyses.

Never underestimate the null hypothesis. Most things don't affect most other things, and p-hacking and its analogues are everywhere. The next time you hear that a single study has found that some obscure vegetable either cures or causes cancer, consider the null hypothesis instead.

Old books have a few things going for them. For one thing, if you come upon an old book, that means someone thought it was worth enough to preserve. The older the book, the stronger the implicit endorsement of centuries of copyists. Hesiod's Works and Days, for example, has won the long battle to hold human attention, and survived while the vast majority of ancient literature has long since passed. We find ourselves in the golden era of old books, at least insofar as availability goes. The Internet Archive and HathiTrust have an enormous number of old books.

They've had a lot of time and seen a lot of things.

They're better than individual polls, which let you fall into the trap of paying the most attention to whatever poll is most interesting. Nate Silver is helpful with this sort of thing.

One of the curious facts about the last two hundred years is that, across a wide (though not universal) set of measurements, things just keep getting better and better. These improvements are enormous and pervasive in ways that people normally underestimate. If you want a very brief introduction to some of this in TED Talk format, you can always take a look at the late Hans Rosling. For books, see Steven Pinker's Enlightenment Now and The Better Angels of Our Nature, Matt Ridley's The Rational Optimist, or Angus Maddison's The World Economy: A Millennial Perspective.

Impossible Food, Beyond Meat, and others are doing some very interesting things with plant-based meat that could very well wind up changing the way food works. I try to check in on how they're doing every so often.

In summer 2019, when I decided to take up messing around with coding after not doing it for about a decade, a younger brother of mine suggested Python 3. It's surprisingly intuitive, and a lot of things are written in it.

Derek Parfit's Reasons and Persons is an excellent book, and works through a number of philosophical questions associated with altruism and morality. It is dense but manageable for a careful reader, even one without formal training in philosophy. Though I certainly don't expect that too many people will slog through it, I'm hopeful it might still leave the world a better place than it might otherwise have been.

The Replication Crisis concerns deep questions about how the public and scientists consume statistical information. Perhaps the best result so far is that replication and meta-analysis have been getting a lot more press lately.

They're good for cooking rice. A good rice cooker -- and I'm not sure there are any bad ones -- runs about $15.^[1] Just go to whatever store you'd buy pots and pans at and get the most basic one. They cook up good rice every time, without any need to fuss about timing or quantity of water or what have you.

SlateStarCodex was one of the most interesting parts of the internet, and currently is kept on ice in more or less the same state it was in in June 2020. It is or was the work of just one person: a pseudonymous psychiatrist named Scott Alexander, who took down his site when, he says, the New York Times threatened to doxx him. After reorganizing his life to avoid the inevitable fallout of being targeted by one of the most powerful media outlets in the world, he now blogs at Astral Codex Ten.

2020

Wikipedia has a helpful article entitled Growth of Photovoltaics, which contains some astounding graphs and charts. From 2000 to 2018, total photovoltaic capacity worldwide doubled approximately nine times, and is now at about 3% of global electricity production, according to the article. If this number is even approximately correct, and it probably is, then we are approximately five doublings away from all the world's electricity being produced by solar cells. Of course, there are some difficult technical challenges to getting there, especially concerning energy storage. For this reason, it is also worthwhile to pay attention to batteries.

Update: 10 February 2023

In 2019, the same Wikipedia chart referred to above says we the world had 630 GWp of capacity for photovoltaic power generation, which climbed to 1185 by 2022. Growth has been close to 25% per annum for 2018-2022.

Statistics are often incomplete, inaccurate, or misleading. Still, if you want to understand something, counting is not a bad place to start. In particular, if you want to understand the world, understanding the concept of standard deviation can help quite a bit.

The Total Fertility Rate (TFR) is a number which estimates how many children the average woman would have given current fertility rates extended forward indefinitely. If you just think of it as "children per woman" your intuition will be roughly correct for most uses. Global TFR has fallen enormously in recent decades (chart here), and that has all sorts of implications for gender roles, the economy, government, crime rates, and the environment.

Wikipedia is the closest thing the world has to a single standardized set of facts about everything. It still has a large number of issues in the content, but there is nothing out there which we could call "like Wikipedia, but better". There's been a long-running debate about whether it is in decline. It is in some ways a dysfunctional place for editors, and it features way too much silly argument in the creation process, with bad actors given far too much time to cause trouble on the site. On the other hand, it's still the leading encyclopedia. It'll be worth watching how things turn out for it, and whether any other general reference tool surpasses it.

This page is released under the CC0 1.0 license.