jdilla.xyz

It’s a boy!

2023-04-20

Things I learned today, via Our World in Data: sex ratios at birth are slightly biased towards boy babies for biological reasons. The “natural” sex ratio is about 105 boys born for every 100 girls.

Why is this? Miscarriage rates are slightly higher for female babies than for male babies, although this varies by phase of the pregnancy.

So the next time you’re guessing a baby’s sex, the odds are in your favor if you guess it’s a boy!

Trends for the next decade

2023-03-23

Trends for the next decade For a lot of reasons, as I survey the landscape of technology, I wonder about the trends and opportunities that will shape the next decade. I remember clearly seeing the iPhone in 2007 and being aware that mobile devices were going to be a big deal for a long time.

So what trends rise to this level now? I see four that I would bet on (although I’ve stopped short of crafting falsifiable bets for them… maybe later).

  • Electrification of everything. This trend has two sides: supply of electricity and consumption of electricity. Global warming, of course, is the macro driver of this trend, but electric cars are the most important thing driving it right now. On the supply side, they create demand that pulls in new technology for supplying electricity and scale. On the consumption side, they are driving electrification technologies (batteries, but also others) that will be reused for other use cases. We are on a path towards carbon emitting fuels being antiquated / confined to niche use cases and electric being the default power source for everything. This is the trend I’m most intellectually confident of.

  • Generative AI / Large Language Models. As someone who has now become a daily user of ChatGPT, Bing, and GPT Playground, I’m feel most strongly that this is going to be impactful. I see in my own work places where I’m able to get certain tasks done in a fraction of the amount of time that it would’ve taken previously. And I’m willing to bet that every consumer app from this point forward is going to be built with Generative AI / LLM capabilities in mind, in the same way that apps built after 2007 began to assume mobile phone capabilities. This is the trend I believe in the most (emotional confidence?)

  • Distributed work. I’m choosing my words carefully here and specifically not saying remote work because it seems clear to me that people and companies don’t really want to be entirely remote. The end of the pandemic makes it seem like this is over, but in reality it has just begun because the equilibrium has shifted away from the pre-pandemic norm of 5 days in the office, but a new equilibrium hasn’t fully formed yet [0]. The more time goes on, the weirder this is going to get, not just for companies, who have to figure out how to offer jobs / benefits / compliance ~globally or nationally, but cities and regions that are going to see patterns of living change substantially.

  • Decarbonization / carbon removal technologies. This is the trend I’m least confident of, however it is obvious that the only way we’re going to meet net zero carbon on the time horizon that we need to is by getting good at pulling carbon out of the atmosphere (and avoiding using it in the first place). I’m distinguishing this from the electrification like this: switching from gas heat to a heat pump is electrification; switching from a coal power plant to a nuclear power plant or coming up with clever ways to remove carbon from the atmosphere is decarbonization. It’s possible that the energy in this space dissipates, but I think that as we move forward in time it will become ever clearer that having a capacity for pulling carbon out of the area is a necessity and the cost of doing this will put pressure on companies to stop putting it there in the first place.

Thought about it, but decided against it.

I see potential in each of these but couldn't quite bring myself to commit to them.

  • Restoring of manufacturing. I just don’t totally believe that American society is going to bring non-defense critical manufacturing back, although I’d love to be wrong.
  • China’s demographic decline. I think this is an underrated trend, but the impacts are more than 10 years away.
  • The demographic ascent of India. Same rationale as above, but less confidence.
  • Loneliness. As more and more friction gets pulled out of substitutes for human interaction like video games and Netflix, I think loneliness is going to become the default state for more and more people. I also think the pandemic pulled forward growth in loneliness in the same way it pulled forward growth in remote work and online grocery ordering. This is the one I got closest to including in the main list.
  • Decline in religiosity in the United States Mainstream Protestant religions seem poised to die out over the next ~20 years. What happens next? Will they consolidate? Or will new movements substitute for traditional religious (this is one way I think about Qanon).

[0]: It seems like the emerging equilibrium is something like this: 1 day a week for teams that are more or less local to an area, 3-4 times a year for teams that are distributed across regions; but how much longer will there be a critical mass of people within regions where it is feasible to do a day a week on the office?

15th century Venetian beads found in Alaska

2023-03-02

Not new but new to me; interesting never less. From Smithsonian Mag:

Archaeologists dug the beads up in the late 20th and early 21st centuries. Now, a new study published in the journal American Antiquity asserts that the glass objects are among the oldest European-made items ever discovered in North America.

Per the paper, Michael Kunz of the University of Alaska Museum of the North and Robin Mills of the Bureau of Land Management studied ten glass beads found at three sites along Alaska’s Brooks Range. The researchers used mass spectrometry carbon-dating to analyze trace amounts of twine discovered alongside three of the beads and date the artifacts’ creation to between roughly 1397 and 1488.

The beads provide evidence for overland trade networks reaching Alaska from the East (Siberia).