thomas

1/ I have experienced, and can relate to, some version of Taleb’s conjecture (https://x.com/nntaleb/status/1839063309133226102). A short 🧵

PhD in aviation ∩ climate science | Aviator, diver, ETH maxi, nexialist | Host of /diving /ethfinance /eth-staker /geopolitics /science /singapore /thomas

3/ Knowing what to do with this information (choosing the correct directionality) is *hard*. The market has surprised me more than once with “sell the news” reactions (indicating the prediction was widely shared and priced in), which of course turn to “buy the news” the moment you try to be clever and short it.

For non-binary events, like the Russian aggression of Ukraine, second-order implications are even harder to predict.

4/ Sizing the impact of the news is equally difficult. Announcements that some expected to be momentous (e.g., the ETH ETF getting approved) turned out to be nothingburgers in terms of impact (at least within a reasonable time frame).

Conversely, barely-noticed press releases (like the release of GPT–1 in 2018) should signal a tectonic shift for decades to come, but typically don’t.

5/ Lastly, actually acting on this gut feel is tough as well. We’re all busy and/or lazy in varying degrees, and having the conviction to act decisively is a rare trait.

It’s also viewed as brazen recklessness in hindsight if the bet doesn’t pan out (a typical Gettier problem of moral luck: you end up a visionary for being right or an idiot for being wrong, yet you have no control over the actual outcome from the moment you make the bet). 🎬

2/ I have occasionally had strong conviction that an event would happen that eventually happened. This is an imperfect but close enough analog to having a crystal ball in this gedanken, with the caveat that others may have foreseen the event too, since I’m talking about somewhat predictable events (e.g., Trump winning in 2016) and not complete black swans (e.g., COVID-19).

This is pretty much what Asimov tried to do with psycho-history in Foundation...

Also worth keeping in mind that just predicting an event correctly isn't enough. You need to know both the event and the *positioning of the market* to predict how the market will move