Did CoViD-19 come from a lab? The truth.
A man bets $100,000 that CoViD-19 doesn't come from the lab. Here is why this is a good example of how to find the truth.
The truth?
In post-truth times, isn’t it presumptuous to even speak of truth at all? Of course, we can philosophize about it wonderfully, but I wish for a practical approach.
As an example, today let’s consider the origin of CoViD-19. It is undisputed that it began in the city of Wuhan. The two most common theories are:
Lab leak: The Wuhan Institute of Virology researched coronaviruses and conducted Gain-of-Function experiments, where researchers try to enhance viruses to understand them better. CoViD-19 was created, and then it accidentally escaped from the lab.
Zoonosis: CoViD-19 is very similar to a virus found in bats. Such bats were sold at the Huanan Seafood Wholesale Market in Wuhan. One day, the virus mutated and got transmitted to humans.
Two other theories would be "it was an intentional bioweapon test" and "researchers discovered the virus in animals." I won't go into more detail on those. I just want to clarify that these are not part of the two theories mentioned above.
Manifold has a fairly popular market on this. If you read the comments, you'll find very extreme opinions clashing.
As a layperson, this is, of course, a difficult topic, but shouldn't the experts be able to find a clear answer? Interestingly, this looks quite different from the prediction market above.
One can see that this expert survey is quite clear in favor of Zoonosis, but on the other hand, there are also 5 people here with over 90% certainty who have the opposite opinion.
What is the truth now? Couldn't some fact-checker compile all the arguments and evaluate them properly?
Rootclaim is an organization that does just that. It is sponsored by Saar Wilf, a startup millionaire, and investigates various topics: chemical attacks in Syria, the shooting down of a civilian plane, did Pakistan know where Osama Bin Laden was hiding, and also the origin of CoViD-19.
Result: Definitely a lab leak.
But why should one trust Rootclaim now? They are so confident in their analysis that they bet $100,000 on it. One hundred thousand dollars!
Last year, a challenger appeared: Peter Miller. His analysis concluded: Definitely Zoonosis.
The $100,000 bet
The bet in detail:
Rootclaim and Miller agree on two judges.
There is a debate in three sessions, allowing each to present their arguments thoroughly and answer all questions from the judges.
Each judge makes an individual decision: Zoonosis, lab leak, or undecided.
If both judges determine the same winner, he wins the opponent's $100,000.
Rootclaim had essentially already publicly agreed to participate in advance. For Peter Miller, the question was: am I sure enough that I am right and that I can convince two judges?
I don't know how wealthy Mr. Miller is, but I think I read a comment that the amount was not a trivial sum for him.
What would I bet $100,000 on? I can't think of anything where I am so sure that I would bet so much money (and someone could bet against it).
The result
Peter Miller wins.
You can see in the prediction market that Miller was initially the underdog with winning chances around 30%. But when the first videos of the debate were released in January, it went steeply upwards. Apparently, Miller was rhetorically superior.
The aftermath
As seen in the first diagram above, this bet did not really convince the market. Roughly, there is a change from 75% to 65%, but apparently, the majority still believes in a lab leak. I am clearly Team Zoonosis, but I would also gladly buy some more NO shares at a good price. 😉
Rootclaim's concluding blog post doesn't read like that of a good loser: It's about the format, but their arguments are still correct. Peter Miller is not satisfied with how things went after the debate until the publication.
How can it be that after so much research and discussion among many smart people, the truth is still unclear? Although "unclear" doesn't quite capture it. There are people on both sides who are "very clear" for or against. So, we can't even agree that we don't know enough.
One reason might be that there are scientific publications supporting a lab leak. These are evaluated differently by experts today (see GCRI survey above), but that doesn't stop laypeople from citing them as sources. This comment describes the change as follows: "CGGCGG might have looked engineered one month into covid, but not one year into covid." If I understand the further comment correctly, such a sequence with many Cs and Gs is typical for laboratory genetic manipulations. However, these sequences are not useful for the virus, and evolution quickly removes them. CoViD-19 still has this sequence, and it is not useless. So, it is probably not a lab marker.
Bet for the truth
How do we now get to the truth? The prediction market is anything but clear because 65% is close to 50%. In other words, as of today, there is no consensus. It will probably take some time until the market closes. The end date is currently set for 2040.
So, one must rely on their own convictions and bet accordingly. It doesn't have to be about $100,000. Each of us makes small bets, such as going out without waterproof clothing and betting that it won't rain.
But should we perhaps bet money more often? Gambling has a bad reputation. It can be addictive and ruin people. Correct. But it also has its good sides:
A bet is a tax on bullshit. Many disputes are quite futile and would better be settled through a bet.
Clarity: Anyone who makes a bet is motivated to create clear conditions beforehand. No one wants to argue about phrasing later.
One does not forget: The winner of a bet is motivated to remind the loser. Of course, one can evaluate their forecast for themselves, but someone else probably does it more honestly (though with schadenfreude).
Nothing teaches you better about how the world works than trying to predict it. So, if truth means seeing the world as it really is, then one should bet on how it will be tomorrow.
In this sense, Peter Miller is a wonderful example. Congratulations on a well-deserved victory!
This article was originally published in german.
Some insights into Samotsvety, the forecasting dream team. Must read.
Mantic Monday 2/19/24. I feel like ACT writes these much more regularly now. Maybe a little competition by Marketwise motivates him?
Manifold.love made it into the Time magazine. It seems the article was written unrelated to the … uh … event. Something like a cringe musical but also a nerdy dating show but also assisted gambling. Art. I’ll settle for art.
Probably until next week, truthful readers! 😊
ACX also dissected the debate: https://www.astralcodexten.com/p/practically-a-book-review-rootclaim
His footnote about the general opinion:
> A separate market on the lab leak hypothesis itself shifted less, from about 70% to 60%. This could either be because bettors thought Peter was a great debater but wasn’t actually right, or because most people in this (very large) market didn’t even watch the debate. In general I’m not optimistic about markets with no plausible way of ever being resolved.
A Survey of 50+ Good Judgment’s Superforecasters https://goodjudgment.substack.com/p/superforecasting-the-origins-of-the
> 74% likely to have been caused by natural zoonosis
> 25% likely to have been caused by a biomedical research-related accident