2050 Predictions: Eyes, Xi, Apps, Cars, Traits, Bounties
(Thanks to the folks at SLIME MOLD TIME MOLD for organizing a project of us blogger-friends to make predictions for 2050)
I can state flatly that heavier than air flying machines are impossible. - Lord Kelvin, 1895
By 2005 or so, it will become clear that the Internet's impact on the economy has been no greater than the fax machine's. - Paul Krugman, 1998
I’ve been loath to make public predictions, for all the obvious reasons.[1] But here we are. My confidence in these predictions is around 80% - of the seven, I expect one or two to turn out wrong. The worst predictions seem to be around claims that something important won’t happen, rather than that it will, so I'll avoid making those.[2] With that in mind, here are my predictions for 2050:
1) Most myopia (nearsightedness) will be solved preventatively without invasive intervention
That is to say, it will be prevented environmentally or behaviorally. Vision is such an important and expensive sense (~20% of the cortex), that it seems unlikely evolutionarily that poor vision would be typical. In animals, the limiting factor for visual acuity is generally diffraction through the pupil, mediated by eye size, not focal imprecision. Further, if it was a problem of hitting a specific focal length, you would expect a bell curve, with a similar number of nearsighted and farsighted people. In fact, in the US there are probably ~50 nearsighted adolescents for every farsighted one.
There are strong reasons to believe that nearsightedness can be controlled environmentally. Only 1.2% of rural Nepalese children are myopic compared to 59% of East Asian 17-year olds in Australia.
This is likely caused by excessively looking at, and therefore focusing on, relatively nearby objects. Indoor environments tend to aggravate this tendency and there is evidence that time spent outdoors is effective at preventing myopia.
I predict that this finding will become well-established and widely recognized, resulting in parenting norms that prevent the development of nearsightedness. Allergies are currently undergoing a similar trend, where they are prevented or treated through exposure to allergens rather than pharmaceuticals. As this knowledge spreads among the educated, not needing glasses will become a class and cultural indicator, much the way straight teeth are now, which will drive adoption. It turns out there is similar evidence to believe that malocclusion (crooked teeth) can also be prevented environmentally, but I’m less certain of the time frame.
2) The PRC will move to physically control Taiwan
Xi Jinping is 68 now, and will likely want to serve for another 10, 15, or 20 years. He would like to be the Chinese leader to reclaim Taiwan, and will probably do it in his 3rd or 4th terms (2022-2032) when he is still relatively young.
I say “move” rather than “invade” because there are plausible scenarios that don’t involve invasion or even shots fired.
I also think the US will do nothing meaningful militarily in response to this, much like Crimea. “Over the past decade, in U.S. war games against China, the United States has a nearly perfect record: We have lost almost every single time.”
Yes, I already broke my rule about predicting that important things won’t happen.
3) Killer Apps
A) The killer app of VR will be social chat, rather than learning, games, or simulations
Just like we use the supercomputers in our pockets mostly for communication, and even call them “phones”, humans are social creatures who will use this technology socially. We currently use Zoom/Facetime etc. for this, which is terrible,[3] so there’s nowhere to go but up.
B) The killer app of blockchain has not been invented yet
The very first application of blockchain, digital currency, is currently its most used application. This aside, its present use is limited enough that people like to say crypto is a solution in search of a problem. This may be true, much like how number theory was also long “unsullied by any application”, but now makes crypto as well as the modern internet possible.
In the US for example, the total value of the currency is only a fraction of the total value of assets like land and equities. The available potential is still largely untapped, and it seems uncommon historically for new technology to ship with its killer app. The first personal computers did not have spreadsheets and the first iPhone did not have an app store. I can’t predict what the killer app will be, just that blockchain has space to grow, and grow it shall.[4]
4) Road congestion will be a fraction of early 21st century levels
For the following reasons:
- Cheap self-driving rideshares will increase the average car utilization rate, decreasing total vehicles and therefore peak vehicle numbers.[5]
- Self-driving cars will help flow by improving driving efficiency, leveraging inter-vehicle coordination, and reducing accidents.
- Remote working will reduce commuting overall and increase flexibility to allow for more off-peak driving.[6]
5) Trait agreeableness will measurably increase in rich countries
Now among the global rich, human reproduction is largely determined by the desire to bear children rather than other reproductive or environmental fitness factors. It’s no longer a given that you will have children if you can. Therefore, temperamental preference towards having children is a new meaningful evolutionary pressure.
Trait agreeableness is one of the psychometric factors of the Big 5, associated with being “considerate, kind, generous, trusting and trustworthy”. Agreeable people tend to be motivated by relationships and it’s correlated with more children in women (but not men) and earlier motherhood.
Personality is substantially heritable, with agreeableness at about 40%, so we would expect a selection effect to be reflected in subsequent generations. As this may only be relevant for women, the sexes may not be equally affected. It could also be that agreeable women are just more easily pushed into motherhood, rather than more desirous of it.
Regardless, new trends are causing agreeableness to be selected for and it’s heritable, so the population average will trend upwards. This is almost certainly not the only trend that will result from this new selection effect, just the most predictable.
6) Bounties and inducement prizes will be key to advances in R&D
Alongside investment and grants, bounties and inducement prizes will become a large fraction of R&D incentives and funding.
Inducement prizes were historically more influential: Technology from the Longitude Prizes finally allowed ships to accurately calculate longitude and the Orteig Prize sent Charles Lindbergh across the Atlantic. They’re making a comeback though, last year Elon Musk’s $100M Carbon Removal XPrize became the largest inducement prize in history.
This is part of a trend towards faster, less bureaucratic funding mechanisms. As Paul Graham notes about grantmakers, “the skills required for getting money from them end up being completely different from the skills required to use the money well.” Emergent Ventures, Fast Grants, and ACX Grants are attempts to improve the situation through streamlined applications. The logical conclusion of this streamlining is a bounty; completing the project is the grant application. [7]
Full disclosure: I’m currently working on a free and open platform for bounties and inducement prizes. (The best way to predict the future is to create it. - Peter Drucker)
Prediction is hazardous, especially about the future. - Danish Proverb.
That said, we cut some slack on certain predictions, like when Keynes wrote that his grandchildren would have 15 hour work weeks due to rising productivity. Yeah, we still work a lot, but hours per worker has been steadily decreasing and who knows what the future will hold? He might have just been early.
I also realized that we make public predictions all the time, they’re just not framed as such. When you explain an investment, choose a career, buy a property or move to a city, you are making a public prediction. Writing yet another “predictions” piece doesn’t seem that much worse. ↩︎This is somewhat inevitable though, as anytime you predict that something will happen you are implicitly predicting that something else won’t. ↩︎
We learned this in 2020 when we first tried to have Zoom parties. ↩︎
As of right now it’s looking like this future killer app might run on Ethereum, but I’m calling Ethereum (or whatever) the platform rather than the app itself. ↩︎
Rideshares also price in congestion more than private cars. You must bid for a ride during peak demand, and cost to use the vehicle itself is variable rather than fixed. ↩︎
Average commuter delay dropped by half in 2020 compared to 2019. There were factors other than remote work, but it shows what's possible. ↩︎
You might also say the logical conclusion is Universal Basic Income. ↩︎
Thanks to Uri, Kalid, Kaamya, and Olof for feedback on drafts of this.