PatrickDFarley
This method is interesting to me and I’d like to get into it someday. Personally I keep finding that whenever I decline to write something down, that one thing will come back to bite me a few days later (because I’d forgotten it). Do you find that you’re able to mentally keep track of things better than before, even if they’re just vaguely in the back of your mind?
Why pay mind to what’s correlated with being right, when you have the option of just seeing who’s right?
I’m arguing that being right is the same as “holding greater predictive power”, so any conversation that’s not geared toward “what’s the difference in our predictions?” is not about being right, but rather about something else, like “Do I fit the profile of someone who would be right” / “Am I generally intelligent” / “Am I arguing in good faith” etc.
These things are indeed correlated with being right, but aren’t you risking Goodharting? What does it really mean to “be right” about things? If you’re native to LessWrong you’ll probably answer something like, “to accurately anticipate future sensory experiences”. Isn’t that all you need? Find an opportunity for you and your friend to predict measurably different futures, then see who wins. All the rest is distraction.
And if you predict all the same things, then you have no real disagreement, just semantic differences
Fun to do with names. Patrick—English version of a Latin name, Patricius, which means “noble”, referring to the Roman nobility, which was originally composed of the paterfamiliae, the heads of large families. From pater (father), which is Latin but goes back to proto-indo-european. From proto-indo-european pah which means “to protect/shepherd”
Is this an epistemology?
I have experiences, and some interpretations of those experiences allow me to predict future experiences.
I didn’t say it was the answer to everything. The original phrasing was “more truthful.”
These are tautologies. What is the point you’re getting at?
What would it mean for rationality to be “objectively better”? It depends what the objective is. If your objective is “predictive power,” then by some definitions you are already a rationalist.
Is your issue that predictive power isn’t a good objective, or that there are better methods for prediction than those discussed on this site?
If there existed a paradigm that is more truthful than ‘rationality’ as you have been taught it, how would you even know?
Easy. Predictive power.
It seems like you have strong feelings about rationality without actually knowing what that word means here
I really like that last bit about chronological cycles of increasing S-level to “win against” the current level, until physical reality smacks us in the face and we reset. Let me try something:
(Physically) Hard times create S1 men; S1 men create (physically) good times.
(Physically) Good times create S2 men (because there’s free alpha in manipulating S1); S2 men create (socially) hard times (because now you don’t know whom to trust about S1 issues)
(Socially) hard times create S3 men (because tribalism builds/confirms social trust); S3 men create (socially) good times (you have a whole tribe or church or culture war faction that you trust).
(Socially) good times create S4 men (because there’s free alpha in manipulating S3); S4 men create (physically) hard times (because they’re disconnected from physical reality).
I’m gonna be lazy and say:
If it comes up tails, you get nothing.
If that ^ is a given premise in this hypothetical, then we know for certain it is not a simulation (because in a simulation, after tails, you’d get something). Therefore the probability of receiving a lollipop here is 0 (unless you receive one for a completely unrelated reason)
The next step will be to write a shell app that takes your prompt, gets the gpt response, and uses gpt to check whether the response was a “graceful refusal” response, and if so, it embeds your original prompt into one of these loophole formats, and tries again, until it gets a “not graceful refusal” response, which it then returns back to you. So the user experience is a bot with no content filters.
EY is right, these safety features are trivial
certain number of heads
Do you mean “certain number of wins”? Number of heads is independent of their guesses, and number of correctly-guessed heads is asking a different question than the original experiment
I was going to say that this week marks the end of the Covid posts being majority Covid content.
My ideal future has Zvi posting nationally renowned journalism on all manner of current events, but all articles have the title “Covid <date>: <tagline>” and only the real fans remember why.
now that the masks are mostly gone except for the subway.
I see only about 60% mask compliance in the NYC subway now. I’ve been maskless myself in the subway for months—doing my part in the preference cascade
Related theory is that they’re planning for a more dangerous disease to be released in the future, either accidentally or on purpose, and they feel the need to perfect their zero-[disease] protocol now. They can’t accept a superficial failure with covid because that means accepting a critical failure with the next thing, especially if they’re not so good at making vaccines.
Paschal’s targeted advertising: How can you be against targeted ads when they’re showing you deals that have positive EV for you?
There’s an attention cost with evaluating whether the deal is in fact positive EV. And effective ads will mostly have a higher attention cost—There’s a “valley of difficult choices” where the EV is close to zero. Most ads you see are to the left of the valley: strongly negative-EV deals that you don’t really consider. But more effective targeted ads will move the needle to the right on average, forcing you to pay more attention to all ads because now their expected ability to give you good deals is higher. (so basically, “with the changes in attention cost, it’s not actually positive EV”).
Privacy—most of us value it as a good in itself. We have nothing to hide but we still don’t want to show you. We get a bad feeling knowing that some random uncaring stranger knows specific details about us. (so basically, “with the utility cost in privacy, it’s not actually positive EV”).
Randomized priorities- sure I do actually want the gardening tools, but I was gonna look at that two weeks from now, after my Florida trip. I have to plan my Florida trip right now. - And then I see the ad for exactly the product I want, and my attention is too hijacked to ignore it. The decisions I have to make in the near future are ordered by priority, and it takes some amount of mental effort to enforce that priority. Targeted ads actively fight that order by taking some random thing I want and asking me to make that decision right now. (so basically, “with the increased willpower needed to enforce decision priority, it’s not actually positive EV”).
All of this stands against a backdrop of: It’s actually really easy for a consumer to take initiative to find the product they want. It’s never been easier compare alternate products and get a view of the whole market for something. So this is the era in which we least need companies to take the initiative to find us.
Did I manage to actually convey something meaningful to you or did I just wordcel 5,000 nice-sounding words together? How would you be sure?
I think you can actually judge that by the value/effort balance of the communication.
I see a kind of spectrum between teaching and.. let’s call it meditation (as in “meditate on X”), where both can convey meaningful ideas and concepts, but the latter takes much more effort to get anything useful, and yields more random results.
With teaching, I’m probably getting all the intended ideas on my first interpretation, and then it’s reliably useful to me. With meditation, I have to bring a lot more effort and ideas myself to try to get something meaningful out of it, and I might still be wrong. You can pick any random sentence and ponder it as a koan, free-associating about it until you feel like you learned or realized something useful. But that’ll feel very different from just going on wikipedia and learning something useful.
It is a spectrum though. GPT-3 doesn’t give random sentences, but when I play around with it and occasionally find something “useful,” it feels more like I’m doing the koan thing (more effortful). Reading good blog posts is much less effortful per value added.
Option two is to point out that they’re talking sense now and acting compatibly with life, and one could not reasonably ask for more than that.
I’m mostly in the second camp. The penalty for being late should not be death, so go and sin no more.
Not sure I agree but I’ll have to think about it more.
Simulacrum-2 doesn’t mean “saying false things”—sometimes S2 says true things. When you discover that someone’s lied to you, do you shame them for saying the false thing, or do you shame them for being on S2 toward you? The latter is a kind of meta-lie that they’re always committing, even when they say true things. The meta-lie is “I’m saying this cause it’s true, not primarily to manipulate you”.
In personal relationships, you can forgive and give people chances to stop being S2 toward you. Or you might even tolerate their being on S2, because you’re convinced they really care about you (like an overprotective mom whom you love). But for public figures / “Narrative”, might it be acceptable to just say “they told a meta-lie, they still haven’t explicitly come clean on that meta-lie, so F them no matter what they say in the future, they blew their chance to ever earn my respect.”
I agree, but I’d lump all of that into “Analyze the circumstances that caused it”. Maybe I should’ve included more external examples like these