Sunday at Dark Markets, I tackled a thorny angle of the #tescreal critique:

What exactly is so bad about Rationalism? 

After all, didn't rationality and Enlightenment create great abundance?

But Rationalism isn't really about improving rationality. 🧵

Sunday at Dark Markets, I tackled a thorny angle of the #tescreal critique:

What exactly is so bad about Rationalism? 

After all, didn't rationality and Enlightenment create great abundance?

But Rationalism isn't really about improving rationality. 🧵

https://tinyurl.com/55nmthun

Rationalism is both a specific cluster of organizations around Eliezer Yudkwosky's LessWrong, and a general tendency along the same lines.

The core of that tendency is dismissing all modes of knowledge *other* than pure reason, such as tradition, emotion, and intuition.

👁Dark Markets Newsletter

"SBF and the Tech Utopians," Repeater Books, Fall '25

Crypto OG. X-CoinDesk, Fortune. 
@WestHamgc
. PhD UIowa

'Hardcore' - NYMag

This is fundamentally wrongheaded from the start, for two reasons:

1) Intuition and emotion are powerful tools for improving human decision making. People who ignore intuition make bad decisions.

2) Godel's Incompleteness Theorem = there can be no total axiomatic system.

Rationalists consider reason potentially all-powerful - including (ignoring Godel) the power to predict the future. 

This futurity is in turn one of the pillars of EA, longtermism, and transhumanism. 

Rationalism in essence led Sam Bankman-Fried to steal #FTX customers' money.

That's because Rationalism has a wrinkle: 

The implicit belief that *the Rationalists themselves* are the only reasonable people. Hence they deserve the bulk of funding to "study rationality." 

As reported last week, $5 million in stolen #FTX funds allegedly went to LessWrong.

Confidence about predicting the future translates into an apologetics of fascist violence.

Nick Bostrom and Will MacAskill think impacts like "massacres" in the next "100 or (even 1,000) years" are "little more than tie-breakers" when judging outcomes.

Confidence about predicting the future translates into an apologetics of fascist violence.

Nick Bostrom and Will MacAskill think impacts like "massacres" in the next "100 or (even 1,000) years" are "little more than tie-breakers" when judging outcomes.

https://tinyurl.com/3d9tsfyb

As usual, these guys watched a sci-fi movie (Dune) and learned all the wrong lessons.

As we must ask all too often these days, *What are they giving themselves permission for?*

Even short of their Malthusian impulses, these people's thinking is dangerous because, counterintuively, *they don't believe in markets.*

After all, if you can predict the future, you don't need the "wisdom of the crowd" to allocate resources.

This is how SBF got blown up.

Finally, and even more ironically, Rationalists don't even believe in the "marketplace of ideas." 

They don't participate in the academic career structure or peer-review system that hones philosophical thought ... because they've got that Peter Thiel money.

The result, of course, is that their ideas are comically bad, and treated with disdain by serious thinkers.

But they're not supposed to be serious thoughts: "AI Alignment" is just an ideological framework for maximizing investment in both sides of an entirely fake problem.

The real problem with AI isn't an imaginary future superintelligence; it's what biased sorting algorithms are doing in the present day. 

That can't be performatively "solved" by funneling money into fake "research institutes".  That's why Google had to silence 
Timnit Gebru.
.
fin