I’m pretty sure (epistemic status: Good Judgment Project Superforecaster) the “AI” in the name is pure buzz and the underlying aggregation algorithm is something very simple. If you want to set up some quick group predictions for free, there’s https://tinycast.cultivatelabs.com/ which has a transparent and battle-tested aggregation mechanism (LMSR prediction markets) and doesn’t use catchy buzzwords to market itself. For other styles of aggregation there’s “the original” Good Judgment Inc, a spinoff from GJP which actually ran an aggregation algorithm contest in parallel with the forecaster contest (somehow no “AI” buzz either). They are running a public competition at https://www.gjopen.com/ where anyone can forecast and get scored, but if you want to ask your own questions that’s a bit more expensive than Swarm. Unfortunately there doesn’t seem to be a good survey-style group forecasting platform out in the open. But that’s fine, TinyCast is adequate as long as you read their LMSR algorithm intro.

# dvasya

The books are marketed as “hard” sci-fi but it seems all the “science” (at least in the first book, didn’t read the others) is just mountains of mysticism constructed around statements that can sound “deep” on some superficial level but aren’t at all mysterious, like “three-body systems interacting via central forces are generally unstable” or “you can encode some information into the quantum state of a particle” (yet of course they do contain nuance that’s completely lost on the author, such as “what if two of the particles are heavy and much closer to each other than to the third?”, or “which basis do you want to measure the state of your particle in?”). Compare to the Puppeteers’ homeworld from the Ringworld series (yes, cheesy, but still...)

(epistemic status: physicist, do simulations for a living)

Our long-term thermodynamic model Pn is less accurate than a simulation

I think it would be fair to say that the Boltzmann distribution and your instantiation of the system contain not more/less but _different kinds of_ information.

Your simulation (assume infinite precision for simplicity) is just one instantiation of a trajectory of your system. There’s nothing stochastic about it, it’s merely an internally-consistent static set of configurations, connected to each other by deterministic equations of motion.

The Boltzmann distribution is [the mathematical limit of] the distribution that you will be sampling from if you evolve your system, under a certain set of conditions (which are generally very good approximations to a very wide variety of physical systems). Boltzmann tells you how likely you would be to encounter a specific configuration in a run that satisfies those conditions.

I suppose you could say that the Boltzmann distribution is less *precise* in the sense that it doesn’t give you a definite Boolean answer whether a certain configuration will be visited in a given run. On the other hand a finite number of runs is necessarily less *accurate* viewed as a sampling of the system’s configurational space.

we can’t run simulations for a long time, so we have to make do with the Boltzmann distribution

...and on the third hand, usually even for a simple system like a few-atom molecule the dimensionality of the configurational space is so enormous anyway that you have to resort to some form of sampling (propagation of equations of motion is one option) in order to calculate your partition function (the normalizing factor in the Boltzmann distribution). Yes that’s right, the Boltzmann distribution is actually *terribly expensive* to compute for even relatively simple systems!

Hope these clarifications of your metaphor also help refine the chess part of your dichotomy! :)

There’s nothing magical about reversing particle speeds. For entropy to decrease to the original value you would have to know and be able to change the speeds with perfect precision, which is of course meaningless in physics. If you get it even the tiniest bit off you might expect _some_ entropy decrease for a while but inevitably the system will go “off track” (in classical chaos the time it’s going to take is only logarithmic in your precision) and onto a different increasing-entropy trajectory.

Jaynes’ 1957 paper has a nice formal explanation of entropy vs. velocity reversal.

design the AI in such a way that it can create agents, but only

This sort of argument would be much more valuable if accompanied by a specific recipe of how to do it, or at least a proof that one must exist. Why worry about AI designing agents, why not just “design it in such a way” that it’s already Friendly!

I agree, it did seem like one of the more-unfinished parts. Still, perhaps a better starting point than nothing at all?

Check the chapter on the A_p distribution in Jaynes’ book.

Losing a typical EA … decreasing ~1000 utilons to ~3.5, so a ~28500% reduction per person lost.

You seem to be exaggerating a bit here: that’s a 99.65% reduction. Hope it’s the only inaccuracy in your estimates!

The main problem with quotes found on the Internet is that everyone immediately believes their authenticity.

-- Vladimir I. Lenin

Here’s another excellent book roughly from the same time: “The Phenomenon of Science” by Valentin F. Turchin (http://pespmc1.vub.ac.be/posbook.html). It starts from largely similar concepts and proceeds through the evolution of the nervous system to language to math to science. I suspect it may be even more AI-relevant than Powers.

Hi shminux. Sorry, just saw your comment. We don’t seem to have a date set for November yet, but let me check with the others. Typically we meet on Saturdays, are you still around on the 22nd? Or we could try Sunday the 16th. Let me know.

The Planning Fallacy explanation makes a lot of sense.

I hope it’s not

*really*at 2AM.

While the situation admittedly is oversimplified, it does seem to have the advantage that anyone can replicate it exactly at a very moderate expense (a two-headed coin will also do, with a minimum amount of caution). In that respect it may actually be more relevant to real world than any vaccine/autism study.

Indeed, every experiment should get a pretty strong p-value (though never exactly 1), but what gets reported is not the actual p but whether it is above .95 (which is an arbitrary threshold proposed once by Fisher who never intended it to play the role it plays in science currently, but merely as a rule of thumb to see if a hypothesis is worth a follow-up at all.) But even the exact p-values refer to only one possible type of error, and the probability of the other is generally

*not*(1-p), much less (1-alpha).

(1) is obvious, of course—in hindsight. However changing your confidence level after the observation is generally advised against. But (2) seems to be confusing Type I and Type II error rates.

On another level, I suppose it can be said that

*of course*they are all biased! But, by the actual two-tailed coin rather than researchers’ prejudice against normal coins.

Treating “>= 95%” as “= 95%” is a reasoning error

Hence my question in another thread: Was that “exactly 95% confidence” or “at least 95% confidence”? However when researchers say “at a 95% confidence level” they typically mean “

*p*< 0.05″, and reporting the actual*p*-values is often even explicitly discouraged (let’s not digress into whether it is justified).Yet

*the*mistake I had in mind (as opposed to other, less relevant, merely “*a*” mistakes) involves Type I and Type II error rates. Just because you are 95% (or more) confident of not making one type of error doesn’t guarantee you an automatic 5% chance of getting the other.

Well, perhaps a bit too simple. Consider this. You set your confidence level at 95% and start throwing a coin. You observe 100 tails out of 100. You publish a report saying “the coin has tails on both sides at a 95% confidence level” because that’s what you chose during design. Then 99 other researchers repeat your experiment with the same coin, arriving at the same 95%-confidence conclusion. But you would expect to see about 5 reports claiming otherwise! The paradox is resolved when somebody comes up with a trick using a mirror to observe both sides of the coin at once, finally concluding that the coin

*is*two-tailed with a 100% confidence.What was the mistake?

How does your choice of threshold (made beforehand) affect your actual data and the information about the actual phenomenon contained therein?

There aren’t that “many” other companies. Talk to KrioRus, I know they explored setting up a cryonics facility in Switzerland at some point.