The post I made a couple of days ago about Twitter’s coverage of the sustained terror assault on the city of Mumbai was by far the most viewed post we’ve made here. It obviously struck a chord. Bits of it ended up on CNN Asia.

In that post, I was responding to the social melee forming around the end of a dirty big fire-hose of re-tweeting, rumour and shocked reaction. The thing I felt most strongly about when the terror was actually unfolding second-by-bloody-second was the triumphal glee that some contributors were expressing that Twitter was ahead of the mainstream media in reporting ‘the event’. In some peoples’ minds the main event was the defeat of CNN by Twitter and the ‘victory’ of social media in general. It was pretty ridiculous even at that stage as a very large proportion of the total tweetage was actually reporting what was happening on television - and then the re-tweets with added conjecture that followed.

A very few people indulged in really ugly displays of aggression and hate and this has, sadly, continued.

Yes, it does sound bad.

Some people say that this is just the way a real conversation works, you just put up with the noise and disparate views, but that’s nonsense. In meatspace you either walk away from the twats or you ask them to keep their voices down. In fact, the only conversations in any way ‘like’ the busiest period of the other night’s hashtag Mumbai event are shouty drunken lock-ins, the 11.45 booze-train burger queue at London Bridge Rail Station on a Thursday night, or a football riot. In any case, at that stage we weren’t looking for ‘conversation’ - it was precisely because Twitter was ahead of the curve and there was little ‘official’ information that we wanted to know so badly what was really going on. At that really critical early stage, working out what was happening was made much more difficult by these morons. It may sound elitist but, clearly, all views are not equal.

And this problem didn’t just affect rubber-necking news-junkies like me, it also made it more difficult to for those people who actually really needed to find emergency numbers, or ask for ask for blood donations. It must have been truly horrible if you were there trying to find out if a loved one was okay. The sheer volume of rubbish tweeting made it more difficult to find useful links. There were plenty of them, but maybe 6 out 10 tweets were just noise.

Sadly, it’s unlikely to be the last time Twitter is used in this way, and so I’ve been wondering if there are ways to introduce a filtered version of the next event-stream on a #Mumbai scale. I’m not suggesting replacing the #[insert event name] search stream - but offering an additional, optional stream that has been ‘filtered’ by a team of trusted curators. In a blog post he made yesterday JP Rangaswami (Confused of Calcutta) put it like this:

Sometimes I think about all this as a giant virtual switchboard manned by volunteers, willing and able to help. We should be thinking about how we can improve all this. How we can set up this virtual switchboard effectively. How we can help quash rumours. How we can take the load off the security and emergency services people.

What we need is a team of international storytellers who we can trust to draw it all together, to verify, tag and reference the best quality external sources, to mix these up with edits from the main #twitstream, and to orchestrate an over-arching ‘telling’ of the story within which we can remain as participants in some way, but with filters. To put it bluntly, you’d get better quality info with all the twats filtered out. Some of it could be automatic - anyone using upper case would be automatically removed to start with.

Who could we trust to do this? People on Twitter with a certain level of reputation? People who have passed what would basically be a crowd-sourced audition? And what about brand-endorsed twitterers? Instead of childishly bashing CNN why don’t we all work together to create a global panel of ‘trusted twitterers’ with them - of people who have passed certain filters? Of course, even to think about doing that would require a new attitude to old media. We’ll have to get over the “CNN is so over, dude…” tendency. We must stop thinking that it’s necessary to continually demonstrate and celebrate the power of social media. And we need to be honest about how it needs to evolve and get better. I don’t feel as optimistic as JP does about the self-corrective power of the crowd. I’m not sure it’s good enough to say that because real-life crowds always result in Chinese whispers we should accept it as an inevitable fact of life on Twitter in all circumstances. Of course, it is inevitable - in the sense that we won’t change human nature - but we can use Twitter in many different ways, and one way to improve it for some specific circumstances would be to engineer a social editorial filter. Don’t we want Twitter to evolve, to mature, to become even better? What we use today is a very immature version of what all media might become - ever-present, real-time/near-real-time, personal, participatory. It can continue to be all of these things *and* we can make it better too.

The solution I’ve been pondering is deliberately technology-lite, and that’s because we need to come up with something now. I reckon you could probably do this using a Tumble-log - most of the effort would be involved in creating a network of trusted curators and agreeing some basic groundrules. Nevertheless, it would also be healthy to consider technological solutions. I found one service on the night that claims to have created an algorithm based on filtering contributions at the beginning of an ‘event’ (see below). I can’t work out how and I can’t see much evidence of it. Apologies if I’m missing something - can anyone tell me how it works? (It’s called Tweetip and you can find it here).

And then there’s the The Dashboard of Doom: a horizontally scrolling mash-up of the leading Mumbai related #hashtags, Flickr-streams and Dipity timelines.

Please drop me a line to let me know what you think about ‘the next time’, and how we might clean the signal up a bit.

I’ll leave you with a selection of some of the more striking screen-grabs I’ve taken from the #Mumbai stream over the last few days. And cheers to my brother Ben, who helped out with this and the last post.

Chances are you didn’t know what to say. You probably didn’t realise at that moment, at the moment you posted your very first Tweet, what the hey Twitter was, how it worked or what it did. Can you remember what you said? Now you don’t have to - simply enter your Twitter user name at My First Tweet and be reminded.

This is what I said at 11:47am on 23 October 2006. God I’m interesting.

When news of the ‘terrorist outrage’ broke yesterday evening several people mailed and messaged me with links to the coverage on Twitter. I was awestruck by the live feeds provided at #Mumbai and others (such as Twitter Grid). Having looked around elsewhere, my initial reaction was that the main old-school news agencies like Reuters, CNN and the BBC just weren’t providing the coverage, in contrast to the truly MASSIVE volume of tweeting going on. But as the evening continued my feelings changed about this, and I started to see and ugly side to Twitter, far from being a crowd-sourced version of the news it was actually an incoherent, rumour-fueled mob operating in a mad echo chamber of tweets, re-tweets and re-re-tweets. During the hour or so I followed on Twitter there were wildly differing estimates of the numbers killed and injured - ranging up to 1,000.

The facts are still unknown, but it was clear that ‘citizen media’ filled the vacuum provided by official news agencies (here again, there were rumours doing the rounds that because of Thanksgiving, most of the official media had gone on holiday - leaving caretaker managers in place. It was suggested at one stage that CNN was being run by the IT team…). So much noise. So little signal. Even if the truest signal was actually coming through Twitter it was so drowned in rumour, personal utterance, revenge and irrelvance as to be incomprehensible. In the flattened world of the Social Web there is clearly no filter on decency or taste. That made tweets like these possible:

One of the most unpleasant things to witness was the general ‘whoop whoop’ the more self-regarding voices of the social mediasphere were giving themselves. For some, it seemed like the social media coverage of the event became the story. The real event for quite a number of people last night was: Twitter 1, CNN 0 - which is utterly sad. Old media became the enemies for many, not the terrorists.

If traditional media agencies behaved like this (openly) they’d find themselves in trouble - but no-one regulates the mob and it answers to no-one.

Then this morning I see the word go round that the terrorists themselves might be using Twitter to find out what the security forces were up to. A hush went round:

Which, once again, brought out the worst in many:

There were no doubt many well-meaning people Twittering. Some on the ground were no doubt using the service to share their personal horror and to connect with the outside world must have been a comfort. But very few were on the ground. Most participants were far away. There needs to be some way of working out who in a situation like this has more authority than someone else. Of course, simply being there isn’t necessarily an indication of authority, but it does provide some context. I’d be interested to know if Twitter helped anyone last night to get hold of the right blood.

Many were simply expressing their horror - that’s fine. But do we want this to be ‘mixed up’ with news. Is it helpful or sensible to mash news up with personal reflections? How do we know what’s true any more when everyone’s voice has the same weight?

Last night scared me. We’re like kids playing with things that we still don’t understand. A human tragedy became “something to follow”.

I did find some interesting links were being posted - to journalist Vinu’s photostream, and I thought that NowPublic provided sensible, coherent coverage from a crowd-sourced base. Mainstream media appears to be using the social stuff to create its own coverage - CNN gave Twitter credit last night, and The NYTimes are actively asking for contributions.

The most touching eyewitness account I found came from a blog called ‘A Night Out In Mumbai’, and tells the story of frightened people helping each other in an hour of need. Definitely worth a read - although be careful as it contains more than 140 characters.

On Saturday I did a presentation at Ruby Manor on using recommendation systems in production featuring our plugin, acts as recommendable (AAR).

This was, without a doubt, the best conference I’ve been too - and the icing on the cake was the leftover £500 behind a students bar afterwards - ginger beers for everyone!

Graham Ashton has done a write up of all the talks and Chris Lowis has converted AAR to use the GNU scientific lib.

I was going to use Slideshare, but they seem to have broken it, so you can download a pdf of the presentation here (video will be up soon). The slides don’t make much sense by themselves though.

Iplayer, Hulu, Vimeo and Youtube have made it manifestly evident that TV is facing a huge challenge from the web, not just for revenues and mindshare but also as an alternative channel to market. But it’s not just a quantitative ‘cheap and many’ channels issue; there’s likely to be a profound qualitative change in how we watch television that threatens the value of things that conventional TV people hold as given: things like scheduling, channel brands and the primacy of television commissioning.

I’ve noticed that when TV execs talk about ipTV they focus on the pipe and the device and assume that the content and the viewing experience will stay much the same (a mass media experience). I came away from The Guardian’s Changing Broadcast summit at the Mayfair Hotel last week with the strong impression that they believe that by pushing event-based TV and ‘combining the creative skills of UK TV production with the ubiquity of the UK digital market’ they can confront the multiple threats they face from digital. 

I think it would be interesting to mull over these questions first:

• How can the value of scheduling and channel brands remain intact when we make decisions about what to watch (whatever, whenever) by what our friends and other people we trust are watching?
• What happens when national barriers fall down (and I can watch Hulu in the UK and BBC iplayer in France)?
• How will a long tail of television (with geographic spread, not just historic) change what I watch and how I decide what to watch and the habits I fall into?
• What will be the impact of more access to shortform have on TV viewing behaviour (just check out five.tv/fwd). 
• When I can respond instantly to what I’m seeing, in lots of different ways to different people or just everybody watching with me, over an interface that’s easy to use, how might my viewing habits and - just as interestingly - TV formats change? (And has the failure of red button created a false sense of security?).

My guess is that across diverse viewing contexts and audience segments some very wide differences in behaviour, formats, channel branding and tv discovery will emerge.

It’s easy for the hangover of a lifetime of past perceptions to shroud us from future reality. It’s interesting that most of the people adopting the comforting approach that nothing will fundamentally change were over the age of 40.  At the end of the day there was a panel of thirtysomethings representing Endemol Digital, RDF Digital, Channel 4, BBC3 and Bebo with a completely different mindset because they started their careers at a time when the old models were already looking tired. Here are some refreshing snippets: 

“Everything done on television is done online simultaneously - commissioning is multiplatform” (BBC3)

“What success is like is really hard to get at” (BBC3)

“We are experimenting to try to find out which models work” (Bebo)

“The web allows us to challenge the old model where we got paid by the cost of what we made, rather than by its value” (RDF)

and most intelligently of all:

“We’re having to learn very fast to keep up with the audience” 

Dave Birss of Unchained Guide was a Creative Director at Poke London before leaving to take on more responsibilities at his own company. He has just re-posted a very interesting post on the future of digital agencies on his blog, originally written a year ago, that is very useful and still relevant - especially because we at Made By many work in the digital world. You can read his entire post here.