Systems are Better than Experience

I saw this tweet:

Experience matters, but having a good system or model of the world is better. Someone who is sufficiently smart can deduce how the underlying system works–be it in economics or finance or anything else that is sufficiently deterministic/non-random–and then optimize for that, by predicting how it unwinds or evolves through time.

This is how it’s possible for people without formal experience to surpass experts. Someone who understands the underlying system that governs money flows will have an advantage at investing and trading, even compared to top economists, traders, or hedge fund managers, whose systems or models are incomplete or wrong. If someone has the superior model, they will always come out ahead. This is also why Jane Street and other top hedge funds prioritize reasoning ability more so than experience, for hiring. And as we saw in 2008, the failure to adapt by instead digging into experience, can have dire consequences.

I remember some guy in 2019 had a thesis about buying FNMA in anticipation of Trump taking them private. Even after Biden , FNMA stock is up huge since then, because Trump came back. Correct thesis and implementation on his part assuming he held. Another individual, Danielle Fong, made huge money with SPY puts just before the pandemic got way worse. This was in early Feb 2020.

An example of how I was ahead of the curve was to short bitcoin as a hedge against tech stocks (I also rotated to BRK.b and WMT to buy on the dips). I mentioned this months ago. AFIK I was the first person to come up with this idea. Three weeks ago I also correctly forecasted that Treasury Secretary Scott Bessent’s dismal of Bitcoin purchases would lead to more selling. I was prepared, having shorted Bitcoin. The price has since fallen 10% in an otherwise flat market. The widespread assumption was the US treasury would buy Bitcoin.

The second system is how all capital is flowing into AI. China’s DeepSeek has led to a global AI arms race. No money is flowing to crypto anymore compared to AI. This is also why Bitcoin has done so poorly in recent months and will keep falling. Trump, being as competitive as he is, does not want to lose this race, so he’s putting crypto on backburner and focusing on AI. You can also see this in the divergence between QQQ and NVDA vs Bitcoin’s price.

Luck? Maybe, but I think it can also be explained by being smarter (she was a child prodigy entered college at 12(!) shown below) and hence inferring the correct system and profiting accordingly. This also agrees with how the smartest people in the world, in the aggregate, have better system or models and tend to be more successful overall at life, especially as seen at the tails.

130-140 is not a high enough IQ or that rare. Try 160+, as hers is, and you will see better results vs the typical ‘LessWrong 140 guy’ (which I think is actually closer to 120-130). Cathie Wood, whose IQ probably maxes out at no more than 125, her ARKK Innovation fund has lagged everything since her fund’s inception in 2015. By comparison, I have beaten the market per-annuum over that same period.

But in other situations, IQ is not as relevant. To beat a dead horse, during the pandemic the credentialed experts oversold the efficacy of masks, lockdowns, and social distancing. In hindsight, there was no correlation between the severity of restrictions vs outcomes. Countries with among the most punitive of restrictions, e.g. China, had many relapses.

Average-IQ non-experts were correct about those measures not working, having correctly intuited that Covid was not as deadly as led on by the media. They were also right about the possibility of a lab leak, which again was dismissed by those experts or labeled as ‘racist’. Or the same for Trump’s ‘liberation day’ tariffs, in which experts were again wrong in defaulting to the worse-case scenario, back in April 2025. I took advantage of the panic by buying the dip in both cases, although I didn’t make anywhere close to as good of returns as her.

Of course, there are exceptions where experts are overwhelmingly right compared to autodidacts or laypeople. For the hard sciences, specially math and physics, where is a precise logical flow and a clear distinction between right or wrong, then intuition, ‘first principles’, or ‘vibes’ doesn’t stand a chance as it does with forecasting, economics, or trading irrespective of IQ. There is no getting around having to learn the material.