while rationality may be the common interest of many causes, the "something new" is unlikely to be a specifically rationality thing. It's more likely to be something that some people find interesting and some people find boring, and so the people split into different taskforces to solve different problems.

The taskforces may be really necessary, and it may be really difficult to admit in near mode.

On some level, it feels wrong to try fragmenting the LW community. I mean, I am so happy that I have found such wonderful people... and now should my next step be to choose a topic that doesn't interest most of them, and focus on that? So that some subgroup will be interested in that, and most will not?

Yes, exactly this. Because trying to make everyone interested means staying on the level of generalities, ignoring the virtue of narrowness. You conduct experiments with specific data, and only generalize later. And yet, focusing on the specific feels like deviating from the topic of this website, which is about rationality in general.

Historically, even Eliezer didn't make everyone on Overcoming Bias happy. There were people who didn't care about quantum physics. Actually, even today some people feel like the Sequences would be better without the quantum physics parts; like they ruin the otherwise good advice on rationality. Quantum physics is just a narrow specific topic; why couldn't Eliezer just leave it out? Well, Eliezer had his reasons, but there is a meta-reason that if you start leaving out specific things for the reason that they are not central to the issue of rationality, and that some people may not be interested in them, then what remains? General pro-"rationality" applause lights? Topics too mathematical to have any obvious connection with everyday life?

On the other hand, it may be a pattern that Eliezer split from Overcoming Bias to follow his own topics; and now Scott is similarly trying to apply rationality to politics on SSC... something like it's easier to focus fully on your mission, when you have your own playground, so you really don't have to care about what other people think about your approach. Like, if you have something to protect, it makes sense to bring it on your own turf, where you can protect it better. Maybe it is necessary that the next big "lesswrongish" thing must happen outside of LW. We already have links to rationality blogs here. Maybe we should embrace that.

However, it seems like a waste of resources if everyone is trying to set up and maintain their own blog, and especially the debating software. Fragmentation of the community: good or bad? The readers of LW are not exactly the same as the readers of SSC. For those who are different, it is better to have the sites separated (so e.g. Scott's readers don't have to worry about LW karma assassination). For those who are the same, it would be more convenient to have the same user interface for everything; the same inbox for replies both on LW and on SSC.

How could we better encourage the creation of the taskforces? Do we even have them, explicitly? (I imagine something like "special interest groups" in Mensa: a group of people with a specific goal, that has a name and explicit membership.) Maybe it is psychologically necessary to make the taskforce function better. (The only examples of LW groups that have a name I could give here would be effective altruists and neoreactionaries, and I already feel bad about the latter example: I think it goes against the usual LW approach to politics.)

Maybe LW should explicitly rebrand as a platform for multiple taskforces. That does not mean that everyone here would have to choose one; it just means that the taskforces would be the officially recognized way of how LW community works. If you want to read and talk, welcome! But if you want to do something important, join an existing taskforce or create a new one!

Also, this is how we could measure the effectivity of multiple rationalist groups. We don't all compete in the area of general rationality; how does one even measure that? But we could have multiple taskforves, and some of them would give impressive results, and others would not. Even if different areas are not directly comparable, there would be at least a difference between getting things done and merely talking about getting things done.

To a degree this could be done even without changing software. We could announce the creation of the taskforces in regular discussion, and perhaps have a convention that when an article is published officially in the name of a taskforce, it will use the name of the taskforce in the title (whether it is an article on LW or a link to an article on a different website). Then, the individual taskforces could publish their plans, results, lessons learned, etc.