Posts

Why do we enjoy music? 2024-05-14T08:29:16.803Z
metachirality's Shortform 2024-03-18T06:19:51.759Z
Is anyone working on formally verified AI toolchains? 2024-03-12T19:36:38.780Z
What's up with psychonetics? 2023-09-16T01:12:05.423Z
Notice your everything 2023-08-08T02:38:39.974Z
External rationality vs. internal rationality 2023-08-02T23:29:59.368Z
An explanation of decision theories 2023-06-01T03:42:05.807Z
What should my college major be if I want to do AI alignment research? 2023-05-25T18:23:43.442Z
Has anyone thought about how to proceed now that AI notkilleveryoneism is becoming more relevant/is approaching the Overton window? 2023-04-05T03:06:30.585Z
Asking for a name for a symptom of rationalization 2023-01-07T18:34:30.364Z

Comments

Comment by metachirality on Habryka's Shortform Feed · 2024-12-08T18:56:05.744Z · LW · GW

It ought to be a top-level post on the EA forum as well.

Comment by metachirality on Algebraic Linguistics · 2024-12-08T18:53:33.501Z · LW · GW

Well that's because it's meant to be quantifying over linear equations. and are not meant to be replaced but and are.

Comment by metachirality on Algebraic Linguistics · 2024-12-07T22:04:01.143Z · LW · GW

i is often used as an index in math, similar to how it is used as an index in for loops.

Comment by metachirality on leogao's Shortform · 2024-11-27T09:07:39.760Z · LW · GW

What would an event optimized for this sort of thing look like?

Comment by metachirality on JargonBot Beta Test · 2024-11-01T18:16:50.065Z · LW · GW

Why not generate it after it's posted publically?

Comment by metachirality on Habryka's Shortform Feed · 2024-10-30T03:34:19.569Z · LW · GW

Aaaa! I'm used to Arial or whatever Windows' default display font is. The larger stroke weight is rather uncomfortable to me.

Comment by metachirality on yams's Shortform · 2024-10-18T16:22:46.869Z · LW · GW

Yarvin was not part of the CCRU. I think Land and Yarvin only became associates post-CCRU.

Comment by metachirality on MichaelDickens's Shortform · 2024-10-02T06:40:19.059Z · LW · GW

Maybe make a post on the EA forum?

Comment by metachirality on Glitch Token Catalog - (Almost) a Full Clear · 2024-09-21T21:45:21.178Z · LW · GW

It seems like if the SCP hypothesis is true, block characters should cause it to act strangely.

Comment by metachirality on Lao Mein's Shortform · 2024-09-14T16:06:59.699Z · LW · GW

Does it not have any sort of metadata telling you where it comes from?

My only guess is that some of it is probably metal lyrics.

Comment by metachirality on Lao Mein's Shortform · 2024-09-14T04:38:27.847Z · LW · GW

Is this an LLM generation or part of the training data?

Comment by metachirality on Meno's Paradox · 2024-08-08T06:40:48.926Z · LW · GW

I don't see how 3 follows.

Comment by metachirality on Is an AI religion justified? · 2024-08-07T18:28:50.316Z · LW · GW

That's because we don't have the intelligence to exterminate ants (without causing more problems.)

On another note, if an artificial superintelligence needed a human for something, it would probably be able to find someone it could convince on the spot, no pre-built religion needed.

Comment by metachirality on Is an AI religion justified? · 2024-08-06T22:32:56.684Z · LW · GW

We have nothing to offer. Anything we can do, an artificial superintelligence can do better, with space and energy and atoms we irritatingly take up.

Comment by metachirality on Is an AI religion justified? · 2024-08-06T16:01:10.129Z · LW · GW

Why would we want to worship AI?

Comment by metachirality on Closed Limelike Curves's Shortform · 2024-07-19T07:04:06.764Z · LW · GW

I think the thing that actually makes people more rational is thinking of them as principles you can apply to your own life rather than abstract notions, which is hard to communicate in a Wikipedia page about Dutch books.

Comment by metachirality on Most smart and skilled people are outside of the EA/rationalist community: an analysis · 2024-07-13T15:19:12.355Z · LW · GW

Emmett Shear might also count, but he might merely be rationalist-adjacent.

Comment by metachirality on Most smart and skilled people are outside of the EA/rationalist community: an analysis · 2024-07-13T04:20:51.092Z · LW · GW

IMO trying the problem yourself before researching it makes you appreciate what other people have already done even more. It's pretty easy to fall victim to hindsight bias if you haven't experienced the difficulty of actually getting anywhere.

Comment by metachirality on quila's Shortform · 2024-07-13T04:14:34.141Z · LW · GW

they figure out planting and then rationally collaborate with each other?

I feel like they would end up converging on the same problems that plague human sociality.

Comment by metachirality on quila's Shortform · 2024-07-13T04:10:31.468Z · LW · GW

I think asociality might prevent the development of altruistic ethics.

Also it's hard to see how an asocial species would develop civilization.

Comment by metachirality on TurnTrout's shortform feed · 2024-07-09T02:07:51.280Z · LW · GW

This reminds me of Moravec's paradox.

Comment by metachirality on The Potential Impossibility of Subjective Death · 2024-07-05T06:39:59.892Z · LW · GW

You should read Greg Egan's excellent novel Permutation City.

Comment by metachirality on yanni's Shortform · 2024-07-04T02:20:15.257Z · LW · GW

I think working on safety roles at capabilities orgs is mostly mutually exclusive with a pause, so I don't think this is that remarkable.

Comment by metachirality on Isomorphisms don't preserve subjective experience... right? · 2024-07-03T16:29:35.018Z · LW · GW

Sorta? Usually the idea is that the presence or absence of hardware determines the anthropic probability of being that conscious process, otherwise you would expect to be some random arbitrary Boltzmann brain-like conscious.

Also this is an immediate corollary of the mathematical universe hypothesis, which says our universe is a mathematical structure.

Comment by metachirality on Sci-Fi books micro-reviews · 2024-06-25T05:38:58.823Z · LW · GW

I feel like you're not giving enough credit to Greg Egan since he came up with all the philosophy himself.

Comment by metachirality on Ilya Sutskever created a new AGI startup · 2024-06-20T01:38:05.213Z · LW · GW

Let's hope not!

Comment by metachirality on I would have shit in that alley, too · 2024-06-18T06:09:54.370Z · LW · GW

I remember going to a city and seeing someone on the subway loudly threatening nonexistent people. I wasn't scared, I just felt bad that in all likelihood, the world had failed this person through no fault of their own.

Comment by metachirality on The 27 papers · 2024-05-30T23:53:10.738Z · LW · GW

I like this format and framing of "90% of what matters" and someone should try doing it with other subjects.

Comment by metachirality on MIRI 2024 Communications Strategy · 2024-05-30T06:14:19.068Z · LW · GW

Decision theory/trade reasons

I think this still means MIRI is correct when it comes to the expected value though

Comment by metachirality on What mistakes has the AI safety movement made? · 2024-05-24T05:52:42.637Z · LW · GW

The thing that got me was Pause AI trying to coalition with people against AI art. I don't really have anything against the idea of a pause but Pause AI seems a bit simulacrum level 2 for me.

Comment by metachirality on davekasten's Shortform · 2024-05-16T07:34:43.791Z · LW · GW

A subpoena for what?

Comment by metachirality on Why do we enjoy music? · 2024-05-14T16:32:54.883Z · LW · GW

I don't think I'm really looking for something like that, since it doesn't touch on the perception of music as much as it does the reasons why we have it.

Comment by metachirality on LessOnline (May 31—June 2, Berkeley, CA) · 2024-05-09T03:37:21.291Z · LW · GW

Isn't TLP's email on his website?

Comment by metachirality on metachirality's Shortform · 2024-05-07T13:58:53.880Z · LW · GW

Sure, I just prefer a native bookmarking function.

Comment by metachirality on metachirality's Shortform · 2024-05-07T05:02:36.607Z · LW · GW

I wish I could bookmark comments/shortform posts.

Comment by metachirality on Some Experiments I'd Like Someone To Try With An Amnestic · 2024-05-04T22:12:01.071Z · LW · GW

You can actually use this to do the sleeping beauty experiment IRL and thereby test SIA vs SSA. Unfortunately you can only get results if you're the one being put under.

Comment by metachirality on Shortform · 2024-05-04T22:08:51.092Z · LW · GW

This sort of begs the question of why we don't observe other companies assassinating whistleblowers.

Comment by metachirality on metachirality's Shortform · 2024-05-01T20:08:10.015Z · LW · GW

I think there should be a way to find the highest rated shortform posts.

Comment by metachirality on David Gross's Shortform · 2024-05-01T05:09:35.680Z · LW · GW

I like to phrase it as "the path to simplicity involves a lot of detours." Yes, Newtonian mechanics doesn't account for the orbit of Mercury but it turned out there was an even simpler, more parsimonious theory, general relativity, waiting for us.

Comment by metachirality on What is the easiest/funnest way to build up a comprehensive understanding of AI and AI Safety? · 2024-04-30T20:05:11.016Z · LW · GW

Vanessa Kosoy has a list specifically for her alignment agenda but is probably applicable to agent foundations in general: https://www.alignmentforum.org/posts/fsGEyCYhqs7AWwdCe/learning-theoretic-agenda-reading-list

Comment by metachirality on avturchin's Shortform · 2024-04-29T19:05:27.044Z · LW · GW

We don't actually know if it's GPT 4.5 for sure. It could be an alternative training run that preceded the current version of ChatGPT 4 or even a different model entirely.

Comment by metachirality on Arjun Panickssery's Shortform · 2024-04-29T16:23:52.891Z · LW · GW

I think it disambiguates by saying it's specifically a crux as in "double crux"

Comment by metachirality on Is being a trans woman (or just low-T) +20 IQ? · 2024-04-25T19:26:12.283Z · LW · GW

Copied from a reply on lukehmiles' short form:

The hypothesis I would immediately come up with is that less traditionally masculine AMAB people are inclined towards less physical pursuits.

If it is related to IQ, however, this is less plausible, although perhaps some sort of selection effect is happening here.

Comment by metachirality on lemonhope's Shortform · 2024-04-24T14:14:07.651Z · LW · GW

The hypothesis I would immediately come up with is that less traditionally masculine AMAB people are inclined towards less physical pursuits.

Comment by metachirality on Motivation gaps: Why so much EA criticism is hostile and lazy · 2024-04-22T12:52:57.835Z · LW · GW

This feels like Scott Alexander could've written something about, and it has the same revelatory quality.

Comment by metachirality on ChatGPT defines 10 concrete terms: generically, for 5- and 11-year-olds, and for a scientist · 2024-04-12T21:32:51.764Z · LW · GW

I assume OP thought that there was some specific place in the training data the LLM was replicating.

Comment by metachirality on ChatGPT defines 10 concrete terms: generically, for 5- and 11-year-olds, and for a scientist · 2024-04-12T15:34:01.616Z · LW · GW

I think that requires labeled data.

It doesn't and the developers don't label the data. The LLM learns that these categories exist during training because they can and it helps minimize the loss function.

Comment by metachirality on ChatGPT defines 10 concrete terms: generically, for 5- and 11-year-olds, and for a scientist · 2024-04-12T04:28:43.507Z · LW · GW

I don't think there are necessarily any specific examples in the training data. LLMs can generalize to text outside of the training distribution.

Comment by metachirality on Any evidence or reason to expect a multiverse / Everett branches? · 2024-04-09T13:56:04.367Z · LW · GW

Another problem is, why should we expect to be in the particles rather than just in the wave function directly? Both MWI and Bohmian mechanics have the wave function, after all. It might be the case that there are particles bouncing around but the branch of the wave function we live in has no relation to the positions of the particles.

Comment by metachirality on LLMs for Alignment Research: a safety priority? · 2024-04-05T16:08:34.649Z · LW · GW

Have you tried just copying and pasting an alignment research paper (or other materials) into a base model (or sufficiently base model-like modes of a model) to see how it completes it?