Monthly Archives: March 2008

Hiding amongst the Penguins

At the weekend I watched my older two kids do something fascinating. They played hide-and-seek. They’re nine and eight so they play hide-and-seek all the time. This time, though, they were playing in Club Penguin, a slick, Disney-owned virtual world for kids. Hide-and-seek isn’t a game offered by Club Penguin, though. There are lots of other games but not hide-and-seek.

So my kids here in the house and another nine year-old friend in his own house five miles away improvised a game in the busy public spaces of Club Penguin. Of course, they didn’t think this to be in the least bit remarkable. I guess that’s digital natives for you.

So, to get started, Billie stood in front of the computer in the kitchen, closed her eyes and counted to twenty while Oliver, at a laptop in the living room and Joseph in his house waddled off into other Club Penguin rooms to hide. Of course, Club Penguin would have allowed Billie to find both of her friends at the click of a mouse via her buddy list – so the kids invented a ‘no buddy list’ rule – and they stuck to it.

The seeker wandered from room to room, peering into the crowded spaces for her friend’s avatars. And since avatars in Club Penguin are all… well… penguins, this was definitely not easy: it was all down to accessories like scarves and hats. Billie eventually found both of her penguin friends and then it was somebody else’s turn. What should I conclude from this observation? I guess kids will treat these environments much as they do the real ones they play in every day.

Wikipedia’s self-awareness

Wikipedia: The Missing Manual. John Broughton (O’Reilly 2008).

John Broughton’s a Wikipedia editor. His new book’s about the hidden machinery of Wikipedia, the tools used by editors and contributors to add, improve and, when necessary, remove Wikipedia content. What emerges is a picture of Wikipedia as a busy, dynamic, pyramid-shaped hierarchy with all the hard work going on in the middle layer—where the editors live. The rules they apply and the tools they use are fascinating and much subtler than the pop media coverage of the project could ever allow.

Broughton’s not interested in defending or explaining Wikipedia (it just doesn’t come up). His book is a how-to (a ‘Missing Manual’). This is as it should be: Wikipedia’s part of the landscape now, not a libertarian fantasy or a half-baked geek experiment. Practical books like this always arrive after the first, experimental, phase and before the later, ubiquitous phase. They’re an important sign. They say: “this phenomenon is real and durable.”

So we learn that what’s coming together at Wikipedia is an important institution with deep roots, a sophisticated social model and a rigorous framework for knowledge. Tools of real scholarly and pedagogic value are growing behind Wikipedia’s facade and, of course, the extraordinary thing about this experiment is that there’s nothing to stop any of us participating. I’m profoundly impressed, for instance, by what I learn about the system’s ‘notability’ process.

Something like 1500 entries are deleted per day, mostly because they fail to meet the collective definition of ‘notable’ (important enough for an encyclopaedia entry). But these deletions is not arbitrary. Many are hotly disputed on the site’s talk pages and just as many are reversed by volunteers (they’re called ‘inclusionists’) who wade in to provide some extra notability for threatened entries: by adding citations, for instance. One of these inclusionists is cult novelist Nicholson Baker, as he explains in this lovely piece from a recent NYRB (also a review of Broughton’s book).

Of course, Wikipedia’s not the answer. It’s not our era’s final position on planetary-scale knowledge sharing. In fact it’s a bit of a mess at the moment and seems to be showing signs of early-onset decadence. But I think this is probably a good thing. The idea that the first large-scale collaborative effort to organise knowledge should be the only one or in some way definitive doesn’t seem right: we should probably have a few goes at this at least.

If Wikipedia were to collapse or even to disappear, like the great library of Alexandria in 48BC, it would pretty soon be replaced—and perhaps by something with an improved contribution model or better governance. We’d miss it (I’ll bet those Alexandrian scholars missed their library too) but, sooner or later, it would be a footnote in a successor work’s index.

What’s fascinating here is the way the institution (is that the right word?) is quite systematically bootstrapping its own authority, growing processes and rules that deliberately and incrementally improve the body of work (no matter what the founder gets up to). This counters the Sunday newspaper orthodoxy, promoted by people who haven’t troubled to look at the way the thing actually works, that Wikipedia has a kind of diminished authority and that it can’t acquire real authority because it’s the product of amateurs. Wikipedia improves because it is self-aware…

I’ve written a piece about Wikipedia for Mark Ellen at The Word which comes out on 8 April.

Comedy voicemail service prospers

Spinvox, the service that translates your voicemails into text messages, has raised a wheelbarrow-load of money from various top-drawer sources for international expansion. I like Spinvox and the company’s continued success shows that it’s clever to hang out where voice and data meet. Machine translation – especially in noisy environments like mobile – is hard so it’ll continue to produce opportunities. Until translation is an embedded function, until every operator offers it as a standard (and invisible) feature of voicemail, Spinvox will prosper.

The thing about Spinvox, though, is that I’m still pretty sure there aren’t any machines. People who’ve visited the firm will tell you that there’s a roomful of… well… people translating those messages. And any user will provide plenty of evidence from their own inbox of human fallibility in the call centre. I keep a small collection of genuinely hilarious Spinvox mistranslations.

From a work contact: “Hi Steve. Beats me in the arse. Please give me a call when you’re available”

From a BBC Manager I know vaguely (and whose name is not John Arthur): “Hi Steve. It’s John Arthur here. I do love you very much, though. Can you please give me a call”

From my wife: “Call ham Steve”

Also from my wife (whose name is Juliet): “Hi Chloey. This is Joey”.

If the heavy lifting at Spinvox really is still being done by people, why do they continue to talk about their awesome software? If the software is steadily taking over (which is the most likely explanation) where are these errors coming from: the software or the people? Speaking for myself, Spinvox’s comedy mistranslations are the main reason I continue to subscribe. I do hope they don’t get that software working too soon.

Pledging allegiance

I like the pledge of allegiance idea. I know it’s a bit uncool and as a ‘liberal’ I should reject the idea as jingoistic or backward but I think Helena Kennedy and the rest of the snooty elite have got their response to this one wrong.

Kennedy on Today was shocking: it was as if the last seven or eight years hadn’t happened. The whole post-war social settlement in Britain was thrown in the air when those British kids blew up dozens of their fellow-citizens on 7/7 and the idea that we can carry on roughly as before is jaded and defeatist – Kennedy’s singing loudly with her fingers in her ears. We should really be ready to try almost anything to strengthen bonds within and between the people and institutions of this country.

The reason I like the pledge of allegiance is probably for the same reason it makes other liberals cringe. It’s an artificial event, an invented marker for accession to citizenship. That’s just what we need: something simple and deliberate that says ‘welcome to your nation’! A moment in time. An unembarrassed celebration with some real emotional content —and maybe a speech you have to learn—that kids can laugh and joke about afterwards but always remember as ‘the day they grew up’ or ‘joined the club’ or whatever.

There’s obviously a risk that such an event could become a laughing stock if it’s inauthentic or too cheesy but I think that’s probably a risk worth taking. I’d like to see a debate about this—and maybe some interesting contributions to the shape and content of the event too.

iPlayer’s economics

iPlayer‘s a huge success. Ashley Highfield says so. The first ‘iPlayer hits’ are emerging (Dr Who, Top Gear). iPlayer audiences are typically 10% of the broadcast numbers, sometimes 500,000 in one day. Within a year we’ll see the first streaming blockbuster – a show whose iPlayer audience exceeds its TV audience.

The corporation used to worry about the cost of streaming – the download iPlayer was, in part, aimed at controlling the Beeb’s bandwidth costs by offloading file distribution to downloaders (some have been surprised to see upload traffic from their PCs while viewing a show).

Greg Dyke was on record (can’t find a reference) as saying he didn’t think it was possible to resolve the economics of a system where each new viewer brings an additional cost, even where that cost tends towards zero. The problem, of course, is that the unit cost doesn’t (can’t) actually reach zero but can only bump along asymptotically – and is likely, in fact, to rise each time a service enhancement comes along (better resolution for instance).

The other problem, of course, lies out there at the network’s edge with the ISPs. Highfield is dismissive. iPlayer traffic is a ‘negligible’ proportion of overall traffic, he says. But, logically, it can’t stay negligible. ISPs are going to carry an increasing share of the burden of delivering the BBC’s streaming traffic to their customers. Tipping points will arrive, quality will suffer. A BBC insider told me the other day that the ISPs represent a huge potential problem for the corporation’s streaming plans. He used some graphic metaphors: cliffs, bullet trains approaching mountains without tunnels, that kind of thing. The $64K question is: can a national broadcaster switch a significant fraction of its content from a free transmission model to a costly per-stream network model without trashing its funding economics or the all-you-can-eat internet access model?

And, more to the point, is anyone at the Beeb thinking about this?