2021 Predictions

For the past two years, my friends have had a tradition of making predictions and scoring them at the end of the year. This year, I decided why not just post my predictions on my blog and score them publicly (with a few redacted)? This will hopefully incentivize me to be more thoughtful about them and also will help direct my endless desire for internet acclaim towards something that has a positive side effect.

Each prediction is written in the format, “#. <prediction>: <confidence level (%)>.” Similar to my friend Eric, whose predictions I also recommend checking out, I’ll be scoring my predictions using $ \log $ scores. Also similar to Eric, I’m happy to score other people’s performance on my predictions. To participate in this, follow the instructions Eric gave on his blog (modified with my own deadline and email):

I’d also like to advertise a challenge for my readers. You can email me with your predictions for a subset of my predictions with your own prediction. I’ll judge your predictions against mine using the logarithmic scoring rule. I have the advantage of having chosen the questions, but you have the advantage of having seen my predictions. In theory this gives you the upper hand: for example, you could put down my probabilities for every event except a few where you think I clearly messed up. Consider this an exercise in second-order knowledge: figuring out how much weight to put on my probabilities despite not knowing my reasoning behind them. Your score will be the sum, over all questions you choose, of your log score minus my log score.1 (So you can guarantee yourself a score of 0 by sending me an empty list.)

Please make your predictions by January 17th at 11:59pm ET and send them to me as a text file where each line has the format “Category: [event #], [probability as a decimal]”. I will throw out any predictions that resolve (or nearly resolve) by January 17th.

I’ll be finalizing my own predictions on January 17th as well, after which I’m not allowed to make changes to them. This means that you have a week to send me suggestions for additional things to make predictions about! In particular, I’m interested in finding good indexes for the rate at which people are moving out of cities like SF and NYC (my initial Googling came up short).

Finally, I want to add an important disclaimer that all predictions are descriptive not prescriptive. Me saying something is probable is entirely independent of how much I want it to happen. In other words, keep your ought out of my is-es. On a more procedural note, all dates should be assumed to refer to 2021 unless otherwise stated. Without further ado, as the eternal great one famously said, “thumbs up, let’s do this!”

Predictions #

Science #

  1. Another baby born from an embryo publicly declared to have been modified with CRISPR (or another gene editing complex): 3%.
  2. Proof that $ \rm{P} = \rm{NP} $ (confirmed by Scott Aaronson): .01%.
  3. Proof that $ \rm{P} \neq \rm{NP} $ (confirmed by Scott Aaronson): .5%.
  4. Language model released that uses 10x the # of parameters used by by GPT-3: 40%.
  5. Highest exact match rate on SQuAD 2.0 $ >92.0 $ on January 14, 2022 (as assessed by this Metaculus question): 60%.
  6. Average world temperature increases relative to 2020 as measured by NASA (https://climate.nasa.gov/vital-signs/global-temperature/): 99%.

Tech #

  1. Waymo self-driving car service open to the public in a city outside of Arizona: 20%.
  2. Another company’s self-driving car service open to the public (anywhere): 10%.
  3. SpaceX starship has a successful test launch in which it lands without major issues (defined as causing enough damage to the ship that it can’t be reused): 60%.
  4. First Tesla Cybertruck delivered to customer in 2021: 45%.
  5. Top price performance (in G3D Mark / \$) of GPU $ \geq 10 $ (as assessed by this Metaculus question): 55%.
  6. The average degree of automation changes for key US professions in between December 2020 and January 2022 by $ >1% $ (as assessed by this Metaculus question): 55%.
  7. $ \geq $1 additional major tech company (FAANG, Twitter, Airbnb, Uber, Lyft, DoorDash, Square, and others I may have forgotten with $ \geq $~5,000 employees at my discretion) with an HQ in the Bay Area announces they’re moving their HQ to either Los Angeles or another state: 40%.
  8. Amazon or Wing drone delivery rolled out to some customers (even if just in beta form) in the US: 5%.
  9. Bitcoin price >$10k: 80%.
  10. Bitcoin price >$20k: 70%.
  11. Bitcoin price >$50k: 20%.
  12. Bitcoin price >$75k: 5%.
  13. Bitcoin price >$100k: 2%.
  14. ETH price >$200: 95%.
  15. ETH price >$500: 60%.
  16. ETH price >$1k: 55%.
  17. ETH price >$2k: 25%.
  18. ETH price >$5k: 10%.

NYC #

  1. Andrew Yang participates in the primary for the NYC mayoral election: 60%.
  2. Conditional on his participation, Andrew Yang wins the primary for the NYC mayoral election: 30%.
  3. Conditional on winning the primary, Andrew Yang wins the NYC mayoral election: 80%.

World #

  1. Number of people in the world living in extreme global poverty falls (as measured by https://worldpoverty.io/map) (currently: 743,336,050): 70%.
  2. China attacks Taiwan: 5%.

COVID #

  1. Enough doses of COVID vaccines to inoculate 100 million people in the US distributed between January 9th and March 31st (as measured by GJP): 10%.
  2. … between April 1st and May 31st (as measured by GJP): 65%.
  3. … between June 1st and July 31st (as measured by GJP): 10%.
  4. … on or after August 1st (as measured by GJP): 15%.
  5. 100 million first doses of COVID vaccines administered in Biden’s first 100 days, i.e. as of April 30th, 2021 (as measured by this tracker or an alternative if this goes away): 45%.
  6. 150 million first doses of COVID vaccine administered by June 1st: 40%.
  7. 1 million first doses of COVID administered in NY as of February 1st, 2021 (as measured by this tracker or an alternative if this goes away): 60%.
  8. 9 million first doses of COVID administered in NY as of June 1st, 2021: 30%.
  9. Fewer than 40 million total cases of COVID-19 in the U.S. will be estimated as of March 31st (as measured by GJP): 4%.
  10. Between 40 and 75 million (inclusive) total cases of COVID-19 in the U.S. will be estimated as of March 31st: 5%.
  11. Between 76 and 128 million (inclusive) total cases of COVID-19 in the U.S. will be estimated as of March 31st: 35%.
  12. Between 129 and 200 million (inclusive) total cases of COVID-19 in the U.S. will be estimated as of March 31st: 50%%.
  13. More than 200 million total cases of COVID-19 in the U.S. will be estimated as of March 31st: 1%.
  14. Moderna, Pfizer, or AstraZeneca produce an updated COVID vaccine targeting a mutation in 2021 (as assessed by this Metaculus question): 55%.

Random #

  1. Scott Alexander starts a substack in 2021: 95%.

Personal #

  1. I write >4 personal blog posts in 2021 (not including this one): 65%.
  2. I perform and write up results from $ \geq $1 self-experiment in 2021: 75%.
  3. I publish $ \geq $1 paper (on ArXiv) this year: 60%.
  4. I weigh <=[REDACTED] pounds as of April 1st: 65%.
  5. I weigh <=[REDACTED] pounds as of April 1st: 55%.