Frame Of Stack

It's not just about arguing whether a cube is better classified as hexagon or square. It's also about noticing impossible colors.

Same here

Right, but I was trying to figure out how, and it seemed like it just didn't fail. But, no, I get it now. I was misunderstanding composition-consistency.

Sortition (choosing a random person to select the winner) has both of those properties (plus strategy-proofness). So your summary of the theorem, at least, seems wrong. Is the paper also wrong?

How else are you going to get $i bills to mean anything?

Ok but then you have to contrast with 50% i/o increase. All that cake takes hands to make. There were not so many novelty cakes in 1950. With 50% more labor, there night have been!

I think your original statement treated this kind of thing as an isolated illegitimate effect, people grabbing the reigns via "value loading". What I want to assert is that this kind of thing happens all along the evolutionary path.

Ok so for example, suppose the next century goes well for the human race in terms of survival and maintaining some kind of control over the future. If the next generation has a big honesty fad, this could tip incentives toward honesty in a way which influenced long term values.

I'm not sure that's arguing against anything you think, but it might be.

So my contention is going to be, "niceness" at our tech level is becoming so context sensitive ("niceness"=environmental fit, but environment=other intelligent agents) that a broad variety of fixed points are possible, analogy to sexual selection and peacock tails

And in a broader sense we could say things like "evolution doesn't necessarily favor intelligence, so intelligence is just another subjective direction to take things in"

Intelligence itself is pretty subjective and value-laden when you get down to it (not in an "IQ skeptic" sense, but in a "AIXI isn't quite how I want to define intelligence" sense -- and even AIXI says the prior plays a role in putting a number to intelligence)

So this is an anti-orthogonality-thesis argument? Or do you endorse some form of orthogonality but not so far as "arbitrary intelligence can forward arbitrary goals" and in particular human goals aren't stable?

I would have to talk for at least half an hour (probably hours) to convey why I *really* think it's good, but my quick summary is that intuitionism only gives you the positive side (sometimes called "assertibility") but linear logic gives you the dual ("deniability").

I was *totally* unimpressed by linear logic enthusiasm until I read "linear logic for constructive mathematics" (because I thought I had strong arguments against it).

This was my intuition, but surprisingly doesn't seem to be the dominant suggestion in this thread atm

Make a decision and then do something different from what you decided. What is doing the doing? @MakerOfDecision

Could you explain why naked shorts would accomplish this? (I didn't see this in your Medium write-up of what you learned.)

If that research were to conclude that only cable cars can be safe, so much the worse for rockets. You can't escape e.g. the rocket equation just because you worked in ignorance of it