🏅 Top traders

I'm deciding on NO:

• The lesswrong team has expressed not being interested in a PR, so I don't know if it makes sense to PR (instead of, say creating a tree dialog platform on my on)

• I really really like dialogues I have listened to so far, and they have a peculiar flow and particularity to them that I don't find in "debates" style of arguments that, in my opinion tree dialogues would nudge toward a lot

• (I don't really know why they are so different. Maybe because of "see what the other person is typing as they type" that incentivize you to to tell your own story rather than direct back-and-forth? Could be a culture thing as well)

• Also, I think porting manifold and some quick assistant project is a bigger priority for me right now

Overall, it doesn't seem like tree dialogs would be as productive as what I first thought. I already noted that many of my discussions improved a lot when calling instead of just writing back and forth (especially huge block of text style), and I think this linear flow does a lot for it. Will report with my friends and see if they're still interested in having dialogues

@epiphanie_gedeon I'm interested to hear more details about what feels good about the current dialogues. Also you specified "listen to", and doublechecking you mean "listen to a la the automated AI narration", and if so curious how that contrasts with the reading experience (or whether you just meant "consuming, somehow")

@RaymondArnold Sure!

So one thing that really jumps out to me when I think of the lesswrong dialogs:

• They do not feel like debates at all.

  • Some patterns I have had with written exchanges about fundamental discussions (mainly, cooperationism v anti-suffering) is a sort of very slow and trudging process, with lots of definitions, "what do you mean by that", and it often feels like we are talking past each other.

  • In contrast, the lesswrong dialogs feel very smooth, fluid, fast and constructive. I feel like there is a sort of shared understanding and goal.

  • I think this comes in part from the fact that the dialogs when they start are established around "let's talk about X" and less "let me persuade you that X is true", which is a feature of the matchmaking/pre-filter process.

  • Paradoxically, I think the fact that the text format guides you toward focusing on what you are writing more than replying to the other person is a key component of that. This is also why I think a tree dialog would actually not work, because it would be more in a format of "replying to the answer" than a "shared stream of consciousness" kind of thing

• They feel very story-telly

  • One thing I have been surprised about is how much I am learning about each persons in the dialog. Their point of view, their experiences, etc..

  • I like how many different perspectives and point of view I can gain through them. This was already a strong point of LessWrong in my opinion, but dialogs emphazises this part even more (which is why I've found them fascinating since they first came out)

• It still feels like a discussion

  • There's something about seeing how each persons evolve in their own perspective and considerations. Sharing mini-updates, deltas, point of clarification or "I haven't thought of that before"

  • I like the convention I'm starting to see of "this is what I am taking out of this dialogue"

TLDR: I like that it feels like two stream-of-consciousness that sometimes interact with each other.

Overall to me, they allow me to have a glimpse of very different perspective in a way that comes up organically and see how each evolve and develop.

Also you specified "listen to", and doublechecking you mean "listen to a la the automated AI narration", and if so curious how that contrasts with the reading experience

I almost only ever listen to the podcasts, I read way too slowly. For the dialogues, I tend to have to open the page on a screen to check sometimes. A feedback on that is that saying who is speaking is very unhelpful and cuts the flow of listening. It would work better for me with clearly distinct voices and without announcing who is speaking.

Thanks a lot!

We currently don't want that feature, so would be unlikely to accept the PR. In case that affects your enthusiasm!