Tummling, SideWiki, Twitter and the Tragedy of the Comments revisited

Says Marshall Kirkpatrick in Twitter is More Likely to Be Meaningful Than TV

In one of those conversations, Kevin Marks (formerly of Technorati and Google, now at British Telecom) told me the following: he believes that Twitter is more likely to be interesting than television because we opt-in to particular streams of other peoples’ updates that we find interesting. That creates a positive feedback loop that encourages us to contribute something interesting in return and thus the ecosystem trends towards higher quality content. Do you agree with that?

Marks also said this was an advantage that Twitter and other opt-in subscription-stream formats have over things like YouTube comments. What of the “I don’t care what you ate for breakfast” critique of Twitter? Marks says that’s just people who have an antiquated view of what belongs “in public,” based on a time when content had to go through expensive publishing processes before being broadcast to the public and thus had to be unusually important to be worth it.

I had a great conversation about RealTime and attention with Marshall, but I think he has coalesced two separate thoughts of mine into one here, in an interesting way. I do find Twitter more interesting than TV, but I realise that may not be a common view.

The first point I was making was that 'realtime' is a mistaken emphasis - what is really interesting is the interplay between the formerly required-realtime technologies like radio/TV and telephony that are now able to be buffered, and the formerly delayed response media like writing, blogging, emailing that are now moving to lower-latency modes. I discussed this in The Flow Past Web

My second one was that the other thing that Twitter makes obvious is the value of semi-overlapping publics - that we all see a different web, and that the default assumption that everyone should read every comment on a forum is an idea that fails at scale too, as one troll or disruptive person can spoil everyone's reading - the Tragedy of the Comments.

Twitter's 'Following' model is powerful here for both its first-order and second-order effects.

The first order effect is that by default we see interesting and friendly comments from people we have chosen to follow, which makes us more likely to want to read on. That people favour and retweet and repeat what they find interesting helps us expand our circles of trust outward to new people.

The second-order effect is that as what we see is mostly interesting, funny, polite and so on, we respond in that vein too (assuming that is what we are reading; certainly there can be self-reinforcing intolerance too, but it is more contained).

Conversely, it is possible to have intelligent and thoughtful conversations in a public, read-everything space too, but for this to work there needs to be someone there setting the tone and establishing the norm - being a Tummler. This week Heather Gold, Deb Schultz and I piloted a show on Leo Laporte's podcast network called Tummel Talk about this important skill and phenomenon, with Jerry Michalski as our first guest. We'll be talking about the idea some more on Social Media Hour with Cathy Brooks on Tuesday 29th September

The skill of Tummling is important, and we need to hold it in mind as we build social tools on the web. Which brings me back to Google SideWiki.

At it's heart, SideWiki is yet another blogging tool, where the blogposts happen to be hosted on your Google profile page. However, as it is deployed inside Google Toolbar, you can see the posts attached to the pages that they are written about as you browse to them.

Google attempts to show the 'most important' comments first, using a combination of voting and other ranking algorithms, but it is still attempting to show everyone the same comment ordering, not taking personal 'following' into account. For SideWiki to succeed, I think this will need to change.

Sidewiki does another interesting thing - it matches comments to the same words elsewhere on the web. For example, my comment on Douglas Adams excellent 1999 piece also shows up in SideWiki on JP Rangiswami's blog where he quotes Douglas Adams too.

This hints at a greater possibility for SideWiki - to weave the web together by better by showing commentary across the web from all places that quote and cite each other, correlating by textual quotation and adding annotated links to the commentary from people we trust most.

This is a way Google could use it's scale of indexing to weave a better web for us to read, through our own chosen trusted sources, rather than funneling commentary into being hosted on its own pages.
(original Google Sidewiki comment)