One problem with current discussion groups is that they naturally gravitate towards a single world view.

When people in a community find a double crux which ends up in statements that currently can't collect evidence for, what should the community do?

If the community tries to stick as one then discussion will be dominated by X vs not X. Both sides will be frustrated by the inability to progress or elaborate their view point. People will likely stop participating in discussions that they don't think are fruitful. This will have a network effect pushing the conversation in one direction.

It is important to keep discussion around not X inside the community as there long as there is some chance. So we need a community structure that can keep them around, but out of the way of the people that believe X. There should be a high level view of what is going on in both the X and the not-X sub-communities so that strategies can be decided upon.

To this end we should try the following structure. An email list is created for a certain question - E.g. How can we create clean energy for our future? When an un-provable statement is found that email list should split it two.

In the sub-group you are not allowed to question the un-provable statement. You must take it as given if you want to discuss there. Any disagreement must be done in the original group.

All the topics subject and a summaries posted in both the sub-lists will get posted to the level above (but not the replies). 

The grand parent list will only get a daily summary of the new topics of the grandchild list, And the great-grand parent list and above will only get numbers of posts happening in the great-grandchild list and below.

So lets take our energy future example further.

Let us say the first double crux found is: "Non-stellar fusion is possible/easy in the next 20 years". The community can agree that if we can find non-stellar fusion in 20 years they should work towards that. But they disagree that it can be found within that time. So the community splits, then splits again (the fusion side splitting between tokamaks and containment fusion, the other side between clean fission and renewables etc etc until there are bits of the community that can work together to explore that solution). 

If you subscribe to the original list you will get discussion of the future of fusion but also summaries of discussion  such"Break-through in storage tech makes renewable future cheap" and one email a day with a list of discussions like "Pebble beds vs Thorium reactors". Finally you'd get to see things like a large number of posts happening on the "Gaseous Uranium Hexaflouride" list which might indicate an imminent breakthrough.

You can move up and down the lists, dependent on whether you are working more at the strategy or the tactical level of things. So you only getting the required information.

This method of organisation of discussion would allow multiple discussions to go on, but not fracture or silo the members of the  community with people that agree with them (unless they ignore the upper layers).

The potential problems with this approach:

  • Disagreements about the double crux to split a discussion.
  • Too many levels would hide the lower level discussions
  • Keeping people to the correct discussion level would be tricky.
  • Moderating very niche/low traffic groups would be tricky.

New Comment
26 comments, sorted by Click to highlight new comments since:

Group performance research indicates that the ability for side groups to nucleate around specific research questions who then bring back their findings to the main group is a key component. Sometimes there will be a negative result: that research isn't useful to the main arc of progress. If too many such projects have negative results it will discourage their formation.

nucleate around specific research questions

Note: around questions. Not about specific answers to questions.

I think Romeo is personally familiar with this research. Are you? If not, we are playing a game of telephone: Romeo interprets research papers, then you correct Romeo based on your interpretation of Romeo.

For the purpose of this discussion, the distinction between questions and answers may be semantic. Any 'answer' can be rephrased as a 'question': "Is [answer] the solution to [problem]?"

the distinction between questions and answers may be semantic

Not in this case, I don't think so.

[-][anonymous]00

Based on what evidence?

[This comment is no longer endorsed by its author]Reply

Let us say the first double crux found is: "Non-stellar fusion is possible/easy in the next 20 years". The community can agree that if we can find non-stellar fusion in 20 years they should work towards that. But they disagree that it can be found within that time. So the community splits, then splits again (the fusion side splitting between tokamaks and containment fusion, the other side between clean fission and renewables etc etc until there are bits of the community that can work together to explore that solution).

If we disagree with "is it possible" the next step is to get the credence. There no good reason to be binary.

If you had infinite capabilities and could explore all possibilities then I agree.

It seems though that people do a modicum of research then decide fusion is more credible and work fully on that. They never fully explore the non-fusion possibilities.

The parent of two lists can be seen as an ongoing process of assigning creedence to each side of the disagreement.

It seems though that people do a modicum of research then decide fusion is more credible and work fully on that. They never fully explore the non-fusion possibilities.

No. If I think that the possibility of commercial fusion within the next two decades is 60% I can still fully explore non-fusion possibilities.

No. If I think that the possibility of commercial fusion within the next two decades is 60% I can still fully explore non-fusion possibilities.

I highly doubt anyone has a full view of the entire set of research into the future of energy. Have you been keeping up with LENR . I've not been because I don't think it is very worthwhile. But I would like it to be (a small) part of the view of any energy strategy.

And even if you have, do you know of discussion forums that can accept both classical fusion and strange stuff like LENR and keep working?

YCombinator funds solar startup's, nuclear fission and nuclear fusion.

It's doesn't fund LENR but LENR has it's questionable history. It's likely that

The Breakthrough Energy Coalition by Gates and other billionaire's suggests 6 possible paths.

Have you been keeping up with LENR . I've not been because I don't think it is very worthwhile. But I would like it to be (a small) part of the view of any energy strategy.

There are multiple ways to think about funding LENR or ESP research. But I don't think making a subforum about "what would be if LENR is real" is the best way to go about it.

But I don't think making a subforum about "what would be if LENR is real" is the best way to go about it.

If some people within the community of "people of trying to solve the energy problem" sincerely thought it was worthwhile to spend some time and energy to try and replicate it (or do some other experiments with deuterium in palladium) there should be space for them to discuss that replication attempt. So that people who come later can see it.

I don't think there should be a space for people to rail against big energy or have conspiracy theories without concrete attempts to replicate it.

There are multiple ways to think about funding LENR or ESP research.

This is not about just deciding whether to fund things. At the lowest leaves of the forums would contain the discussion of the research in the trenches. Having a paper trail of all the discussion from the highest strategy to mundane implementation details seems valuable. Managing that without partitioning it in some fashion seems hard. Partitioning on fundamental disagreements would allow people to pick the partitions relevant to their current research interests. So a researcher in bacterial biofuels could pick a high level forum for summaries but keep an eye on mid-level water management strategies (as biofuels require significant water sources) and be actively involved in the "bacterial bio fuels as the most likely source of clean energy" forum. They wouldn't have to be involved in the discussions between the pebble bed reactor people and the thorium people.

I realize that some of my previous comments weren't as good as they could have been. Specifically about the person going into fusion and forgetting about everything else. It was an exaggeration. I'll try and put more thoughts into my comments in future.

I dream of a more pluralistic intelligence based existential risk discussion group and community, where biologists can talk sensibly about brain extension (and its failure modes) without wading through decision theory.

In practice I am exposed to discussion of LENR by LW + the surrounding community.

Metaculus for example has the threads: http://www.metaculus.com/questions/65/will-radical-new-low-energy-nuclear-reaction-technologies-prove-effective-before-2019/ http://www.metaculus.com/questions/18/will-rossis-1mw-e-cat-tests-lead-to-continued-significant-financial-investment/

I don't think there should be a space for people to rail against big energy or have conspiracy theories without concrete attempts to replicate it.

You mean the conspiracy theory that Andrea Rossi frauds his investors?

So a researcher in bacterial biofuels could pick a high level forum for summaries but keep an eye on mid-level water management strategies (as biofuels require significant water sources) and be actively involved in the "bacterial bio fuels as the most likely source of clean energy" forum.

I see no reason why he should spend more time in the "bacterial bio fuels as the most likely source of clean energy" then in the "bacterial bio fuels as an addition to solar for the times when the sun doesn't shine" group.

Managing that without partitioning it in some fashion seems hard.

You argue for more than just "partitioning it in some fashion". Reddit has partition with subreddits. StackExchange has partition with subsites and then tags. Arbital's forum is also supposed to have partition by tag.

You mean the conspiracy theory that Andrea Rossi frauds his investors?

What?... no Where did that come from? I meant the more typical stuff you might find in a LENR forums about big oil companies not wanting you to find out the truth.

Having a subforum for debunking rossi might be interesting if significant amounts of money was going to him or significant attention.

You argue for more than just "partitioning it in some fashion". Reddit has partition with subreddits. StackExchange has partition with subsites and then tags. Arbital's forum is also supposed to have partition by tag.

True. I think complete partitioning is bad. You split the commons. My suggestion was a way of making partial partitioning work by splitting around un-productive questions but still keeping some dialog possible.

What?... no Where did that come from?

The idea of a conspiracy theory is that people secretly plot. The perspective I got through online discussions was that Rossi plots together with other people to fake results and thus got his investors money.

I was using a broader meaning of the words

In general you have forums that engage in that kind of reasoning but they are mostly not forum where anything productive comes out.

Which is why I said:

I don't think there should be a space for people to rail against big energy or have conspiracy theories without concrete attempts to replicate it.

Claim 3: Splitting a discussion list and posting summaries of the sub lists to the parent will allow discussion to continue whilst still providing an overall view

I disagree. Having two bubbles sitting there stewing in their own juices is not discussion and posting summaries up-list won't help because the people from the two bubbles are still not talking to each other. For a fruitful discussion you need cross-pollination, that is, sex, that is, a certain amount of fuckery. Abstinence, of course, is the ultimate contraceptive so by keeping the two parties separated all that you'll achieve is that in a while the update summaries from both will be identical and will state "We are still right and they are still wrong".

But with a double crux that is currently unfalsifiable there is no or little good discussion to be had. Take likelyhood of hard take off vs soft take off vs no take off. We can get some information from our current world but there is no slam dunk either of the dangerous ways (without a verified theory of intellgence), we can inch it one way or another.

It is worthwhile to think about what we should do before hard take off. It is also worthwhile to think about what we should do during soft takeoff in case we find ourselves in that world.

But with a double crux that is currently unfalsifiable there is no or little good discussion to be had.

I don't think that's true. For example, much of what social sciences do is debating unfalsifiable propositions.

In particular, "unfalsifiable" does not mean that the weight of the evidence couldn't be more on one side than the other. If you are not going to discuss questions for which there is "no slam dunk", what are you going to discuss? Whether water is wet?

It is worthwhile to think about what we should do before hard take off. It is also worthwhile to think about what we should do during soft takeoff

Sure, but why is that a problem? Discussing conditional scenarios is a very commonplace activity.

Claim 2: It is valuable to keep both view points on un-provable statements around so that strategic decisions can be made with information from both sides

Provided, of course, that someone needs to and is capable of making a "strategic decision".

Claim 1: Discussion groups tend towards a single view point on currently un-provable statements due to network effects.

I think there are two usual outcomes. One, as you say, involves pushing the opponents out. But there is also the other one where you just plant a mulberry bush in the middle of the forum.

Needless to say I would be interested in this for the future of intelligence.