Filip Sondej

Karma: 197

Filip Sondej 8 Apr 2024 11:30 UTC
11 points
1 ∶ 0
on: EA Forum feature suggestion thread
Expose “voting tribes” in the comments.

We could run a similar algorithm to X’s Community Notes. Then, on contentious topics, we could easily see the main axis of disagreement. We could also have a comment sorting option, that upranks those upvoted by people from both sides of the disagreement.

See this thread for discussion and corresponding post for algorithm description.

Filip Sondej 23 Aug 2023 17:52 UTC
4 points
0 ∶ 0
in reply to: FlorianH’s comment on: New cooperation mechanism—quadratic funding without a matching pool

I would not say my original version of Mutual Matching is in every sense more general.

Ok, I think you’re right, it’s not strictly more general.

QF mainly deserves a name of information eliciting … as … asymptotically all money has to come from the central funder if there are many ‘small’ donors

Agreed!

ability to set (or know) her monotonously increasing contribution directly as a function of the leverage, which I think is really is a core criterion for an effective ‘leverage increasing’

Yeah, that is a crucial component, but I think we need not only that, but also some natural way in which the donation saturates. Because when you keep funding some project, at some point the marginal utility from further funding decreases. (I’m not sure how original Mutual Matching deals with that.) I think Andrew Critch’s S-process deals with that very elegantly, and it would be nice to take inspiration from it. (In my method here, the individual donations saturate into some maximal personal limit, which I think is nice, but is not quite the same as the full project pot saturating.)

Filip Sondej 23 Aug 2023 11:01 UTC
2 points
0 ∶ 0
in reply to: FlorianH’s comment on: New cooperation mechanism—quadratic funding without a matching pool
Oh yeah! They are identical in spirit, but a bit different in implementation. So they will have different results sometimes.

It would be nice to test both of them in some real world setting.

Edit: it seems that your method is more general? And that maybe you could set the curves in a way that the matching works the same as described here. So the method in this post maybe could be seen as a specific case of Mutual Matching, aiming to have some particular nice properties.

Filip Sondej 4 Nov 2022 14:27 UTC
1 point
0 ∶ 0
in reply to: Ariel G.’s comment on: Mind is uncountable
I tend to think that qualia are bound together when they causally act as one. So if left and right are highly integrated (act as one), they aren’t separate experiences. So here I agree with IIT.

Ah, this story is great. In general Egan’s stuff is awesome. If I remember correctly the story was more about personhood (the memories and dispositions etc.) rather than separate experiences (which would require some processes to run separately in parallel). I think it’s an important distinction to make, as experience is fundamentally real, while personhood (or “continuous self”) is more of a thing we assign to systems—a useful fiction (just like money, or democracy is a useful fiction).

There is also a feeling of being yourself, but that’s a different thing than pure experience, and different than assigned personhood. For example there are cases (meditation and psychedelic trips) when the feeling of being someone disappears, but experience remains.

I like that soul swapping argument :D

In the case of split brain, both of the experiences would feel to be me.

Filip Sondej 4 Nov 2022 13:37 UTC
1 point
0 ∶ 0
in reply to: Emrik’s comment on: EA Forum feature suggestion thread
Ah, no, I just read the report of results on Wikipedia (that’s how they worded it). Hm, it’s strange if that’s not in the paper.

Filip Sondej 1 Nov 2022 11:56 UTC
1 point
0 ∶ 0
in reply to: Sharang Phadke’s comment on: EA Forum feature suggestion thread
I don’t suspect it to be that bad. More like some noise added to each post’s score, and some posts not getting enough attention because of that.

In the reddit experiment single upvotes caused posts to have 25% higher mean score later (this effect was present in all parts of the distribution).

But the effect size was very dependent on the topic, so I’m curious how would that turn out for EA Forum.

Filip Sondej 1 Nov 2022 11:33 UTC
1 point
0 ∶ 0
in reply to: Sharang Phadke’s comment on: EA Forum feature suggestion thread
Yeah, good point. It may be mostly redundant.

Filip Sondej 1 Nov 2022 11:30 UTC
1 point
0 ∶ 0
in reply to: Sharang Phadke’s comment on: EA Forum feature suggestion thread
Oh great! I didn’t know about some of them.

Still, the main thing I had in mind was to embed some custom interactive stuff.

Implementing it as iframe support, would be the most general, and you would solve all the possible “embed X” suggestions at once. So it seams to be the most efficient approach.

Filip Sondej 29 Oct 2022 12:33 UTC
3 points
0 ∶ 1
on: EA Forum feature suggestion thread
Embed iframes

Some use cases:
- interactive visualizations
- videos
- airtables (as already suggested)
- polis questions (as already suggested)
- guesstimate models
- and many many more
This feature is very versatile and would solve many things at once.

Filip Sondej 28 Oct 2022 20:31 UTC
3 points
0 ∶ 0
in reply to: Emrik’s comment on: Sort forum posts by: Occlumency (Old & Upvoted)
Great!

I posted it in that thread: link

Feel free to add something there.

Filip Sondej 28 Oct 2022 20:27 UTC
7 points
0 ∶ 0
on: EA Forum feature suggestion thread
Check if information cascades / social influence bias is a problem on EA Forum.

If it is, maybe we could implement Emrik’s idea to counter it, or some similar mechanism.

See here for the explanation of the potential problem.

To test it, we could do an experiment where some bot (or server-side process) randomly upvotes or downvotes new posts. We measure final karma after some fixed time, and see if that single vote snowballed.

relevant discussion
What links here?
- Filip Sondej's comment on Sort forum posts by: Occlumency (Old & Upvoted) by Emrik (28 Oct 2022 20:31 UTC; 3 points)

Filip Sondej 28 Oct 2022 17:00 UTC
7 points
0 ∶ 0
on: Sort forum posts by: Occlumency (Old & Upvoted)
I like that idea about information cascades. We could test how big this effect is on EA Forum, by having some bot who randomly upvotes or downvotes new posts, and measuring the final karma after some time.

There was a similar experiment with reddit (maybe you already know this).

The accumulating herding effect increased the comment’s mean rating by 25% compared to the control group comments (Figure 1C). Positively manipulated comments did receive higher ratings at all parts of the distribution, which means that they were also more likely to collect extremely high scores.

effect was present in the “politics,” “culture and society,” and “business” subreddits, but was not applicable for “economics,” “IT,” “fun,” and “general news”

Why do you think information cascades aren’t significant on EA Forum? (I hope that’s true)
What links here?
- Filip Sondej's comment on EA Forum feature suggestion thread by Aaron Gertler (28 Oct 2022 20:27 UTC; 7 points)

Filip Sondej 28 Oct 2022 15:49 UTC
1 point
0 ∶ 0
on: EA Forum feature suggestion thread
Recommend posts using collaborative filtering (“people who like the same posts as you, also like:”)

MVP could be done quite easily using some of these techniques.

I have some ideas how to do better. If you consider implementing this feature, hit me up to talk!

Filip Sondej 28 Oct 2022 15:39 UTC
8 points
2 ∶ 0
on: EA Forum feature suggestion thread
In-line commenting.
Invisible by default so they don’t distract, but you can easily switch visibility.
So the reader particularly interested in some section could dive into the comments particularly about that section.
Also, as a further feature, you could color code different comment types, like:
- blue (default): just a comment
- yellow: fix suggestion
- brown: link to previous discussion / relevant resources
- red: critique ?
Also see @Emrik’s comment with more rationale.
What links here?
- Filip Sondej's comment on New tool for exploring EA Forum and LessWrong—Tree of Tags by Filip Sondej (28 Oct 2022 13:09 UTC; 1 point)

Filip Sondej 28 Oct 2022 15:25 UTC
1 point
0 ∶ 0
on: EA Forum feature suggestion thread
Bookmark folders.

There should still be the default one, but if you choose you could put the post in some other folder (sorta like youtube does with saving videos to playlists).

It can have many use cases, like:
- prioritizing things to read
- topic specific folders
- maybe even curation, if you could also make those folders public
Right now I’m doing something along these lines, but with an external editor and lists of links, so it’s a bit awkward to use.

Filip Sondej 28 Oct 2022 13:57 UTC
1 point
0 ∶ 0
on: EA Forum feature suggestion thread
An optional reading time indicator, like here: working example (and that tool’s description).

The bar at the right of each post is the reading time indicator. Full bar means 30 min, half bar means 15 min, and so on.

You can find the code that implements that bar here: html, css

The post length is often the deciding factor in whether I want to read something, so it’s nice to have it at a glance. Also I admit I kinda want to incentivize people to write more concise posts :)

Filip Sondej 28 Oct 2022 13:31 UTC
1 point
0 ∶ 0
on: EA Forum feature suggestion thread
Add an option for drafts: “Anyone with a link can read”, but make it really anyone, not only forum users, as it is now.

(Recently I wanted to get feedback from some people who are not on the forum, and I had to copy draft to google doc, and later copy it back, and fix all the footnotes :/ )

Next step (but probably harder), would be to let anyone comment. If they aren’t logged into forum, these comments are anonymous.

Also collaborative editing in markdown mode would be useful.

Filip Sondej 28 Oct 2022 13:09 UTC
1 point
0 ∶ 0
in reply to: Emrik’s comment on: New tool for exploring EA Forum and LessWrong—Tree of Tags
I’d also want to have an option to switch the marks’ visibility (and on default they should be off, to not distract from reading). With that, I wouldn’t even require author approval, it would be more like commenting, but line-specific.

So the reader particularly interested in some section could dive into the comments particularly about that section.

Also, as a further feature, you could color code different comment types, like:
- yellow: fix suggestion
- red: critique
- blue: link to previous discussion / relevant resources
- green: just a comment
FYI: I posted that suggestion on the EA forum feature suggestion thread, and also linked your comment

Filip Sondej 28 Oct 2022 11:28 UTC
4 points
1 ∶ 0
in reply to: Charlie_Guthmann’s comment on: New tool for exploring EA Forum and LessWrong—Tree of Tags
Yeah, I’d love to see some novel curation mechanisms too. I’m a bit scared to introduce money to the mix though. Someone might be tempted to exploit the system, by using bots to fake engagement. Which would be a loss because you could no longer trust vote count as an indicator of value.

Other way to incentivize curation, would be to tip people for great curated lists of posts. (Although I suspect many people would do it even without the money incentive, if the forum made it easy to do.) For example an option to just publish the list of all your strongly upvoted posts, would enable quite effortless curation.

I guess that the main problem isn’t that some posts are objectively undervalued, but rather that it’s hard to find those old posts that are right specifically for you. Posts that are:
- in a topic that interests you
- novel for you
- not so advanced, that you can’t follow them
There is also an ML approach, where you could find some embedding for each post on the forum, and then train your personal classifier to predict which posts you would like.

Or even more simply, a collaborative filtering approach—“people who like what you like, also like this”.

Filip Sondej 28 Oct 2022 11:20 UTC
1 point
0 ∶ 0
on: Database dumps of the EA Forum
Here is some code you can use to get the most essential data from all the forum posts: https://github.com/filyp/tree-of-tags/blob/main/tree_of_tags/forum_queries.py

Also, I used that data to build a topic model of the forum as you mentioned. You can see it here