Inadequate Equilibria by Eleizer Yudkowsky | Hacker News Reader

Inadequate Equilibria by Eleizer Yudkowsky

equilibriabook.com

34 pointslisper8 comments

Eliezer Yudkowsky has an extremely devoted, almost cult-like following, as does the Less Wrong rationality movement.

But I've never seen an appraisal of him or his work from anyone outside the movement that describes its value or lack thereof. Many of his ideas make sense (if you accept all of his theoretical premises), but I'm not sure they're grounded in reality.

Does anyone here know of such an outside appraisal?

Edit: To be clear, I'm talking about the work that is actually unique to the movement, such as the Overcoming Bias blog and other original stuff. Things like Bayesian reasoning are already proven on their own, but also didn't come from Less Wrong originally.

This seems a bit unfair. Outsiders who agree with Yudkowsky are more likely to become insiders.

Anyway (speaking as an insider) Inadequate Equilibria doesn't intersect with Yudkowsky's theories about AI. It's about how to tell the differences between the circumstances where you should trust your instincts vs those where you should go with the crowd.

This seems a bit unfair. Outsiders who agree with Yudkowsky are more likely to become insiders.

Any cult could say that, however. One question I'd ask, is how likely is it for outsiders who partially agree with Yudkowsky to become insiders? Quite often, it's much less the "leader" and much more the mob inside a cult which creates much of the cult-like atmosphere. I remember posting an idea in the HPMOR reddit about creating a muggle magic detection device that could be used to game-balance the power of the muggle world vs. the magical world. The knee-jerk reactions, followed by the feeling that I must be a horrible person for doing so made me pause.

It's about how to tell the differences between the circumstances where you should trust your instincts vs those where you should go with the crowd.

So, in part, it's how to tell if you're being suborned by groupthink.

Always keep in mind that scene from The Life of Brian, where the crowd shouts, "We are all individuals!"

The thing to keep in mind when trying to start a movement of rationalists that doesn't take on the qualities of a cult: Ask yourself if you are as smart and self aware as others who have tried the same. Why is it you will succeed where those others have failed?

(Do I think the Less Wrong crowd is a cult? It's a foregone conclusion that any such group has grown an outer rind which has at least some cult-like qualities. The real question is, what is the state of the inside? And if you are inside, how do you know you haven't been suborned by groupthink?)

Look at how do they react to jokes. My observation is that all those destructive cult-like organization cannot laugh at themselves.

My observation is that all those destructive cult-like organization cannot laugh at themselves.

It cuts both ways, actually. In this, as in many things, there is a moderated middle path. There is a kind of rather derisive, malevolent kind of humor which is used by certain adherents to nihilist philosophies. If you look very closely, then you will see that they also "cannot laugh at themselves" -- but on the surface, it looks like they can.

They say you can estimate the maturity of a group's culture by how easy it is to commit heresy.

Less Wrong fans seem fairly mature by that metric, since they don't blacklist their apostates.

The Yudkowskites actually seem one step ahead of academic culture if you're looking at groupthink indicators.

Eliezer: "And then I ran into Hacker News, where accusations [of being a cult leader] were aimed at the site owner, Paul Graham."

http://lesswrong.com/lw/4d/youre_calling_who_a_cult_leader http://lesswrong.com/lw/lv/every_cause_wants_to_be_a_cult/

Please correct spelling Eleizer ->Eliezer

Reply on news.ycombinator.com