[I have to dash to get supper ready, so accept apologies for this rough draft in meantime – somehow the hook of last night is better than "the evening before yesterday"]
This happened
London #12
Basics
May 15th 2012 at
the V&A - http://www.thishappened.org/events/london-12
Next event (put it
in the diary!) – Oct 5th
Background to This
Happened
This Happened has
been going in London since 2007 (with international offshoots) and its ID
is very simple: an evening all about interaction design and technology, based
on three to four 10 minute talks (each with 10 minutes of Q&A). The talks
tend to be about finished projects, "This [cool thing that]
Happened", so the speaker can give a quick overview of what the project
set out to achieve, but with the focus mainly on what came out the other side
and the lessons learnt.
I've been going to
miscellaneous spoken word events in London since 2008 (I worked for
Intelligence Squared and have experienced the gamut of "intelligent"
events in London e.g. British Library, the RSA, Southbank Centre, 5x15, and the
different bits of the University of London), but despite having no background
in interaction design and only a very lay understanding of technology, I'd say
I'm most excited by my trips to This Happened. There's something special about
atmosphere – the joie de vivre with which everyone shares their projects
is infectious and the onus on imaginative side-projects gives the evenings a
pleasing amateurishness which most spoken word events are lacking. I might be
the exception, but I have never been lured to a This Happened event by a known
name or project – rather it's been the expectation that intelligent,
creative people will be introducing one of their past projects in a way that
looks to entice as much discussion in the Q&A as possible. As Alexandra Deschamps-Sonsino (@iotwatch, co-organiser of TH London) put it whilst introducing the event, "if no-one asks any questions, you may as well be listening to
television."
Last night didn't disappoint. There
were four brilliant talks and here's my summary of each.
Haiyan Zhang
(@haiyan and interaction designer at IDEO) was the main driving force behind
Japan Geiger Maps (http://japan.failedrobot.com), a "real-time"
visualisation of geiger counters from across Japan in the aftermath of the
March 2011 earthquake and nuclear catastrophe. In the weeks and months
following the explosion at Fukushima there was a lot of data being circulated
about radiation levels in the vicinity and across Japan, but Haiyan wanted to
turn this into useful information which average people could understand
"at a glance". Using data from Japanese government sensor networks,
the Safecast fixed sensor network and individual (private) geiger counters,
Haiyan mapped the radiation levels on to google, displaying the numerical
measurements (in µSv
per hr) as well as relative strengths (compared to the average public space
geiger reading for Japan) by using a colour code. Despite difficulties in data
accuracy and without any marketing the website attracted 500,000 visits in its first six months.
Possible because:
Internet of Things – lots of geiger counters
(especially on the Safecast network) were outputting live data on the internet.
Crowd-sourcing – citizen scientists in Japan
were keen to contribute as individuals to the larger data set
Lesson:
Turning data into information is a great first
step, but then it's essential not to lose sight of the need to maintain
quality. The accuracy of the readings (especially from individuals) was
hit-and-miss and the representation of "difference from the norm" is
inherently problematic.
Adrian Hon (@AdrianHon and co-founder / CEO of Six to Start, creators of
Zombies Run! (here's a great feature on the app by Tom Chatfield in the Observer) couldn't get much further than the tip of the iceberg in his talk on Perplex
City, "a worldwide alternate reality game/collective puzzle
card game/treasure hunt that ran from 2005 to
2007, involving a £100,000 prize and 512 puzzle cards containing clues and
rabbitholes
to an extensive online world." If I'd
noticed one of these puzzle cards in Borders, I'd probably have failed at the
first hurdle, but it sounds like if you were a whizz at cryptic crosswords and
liked intricate stories (the narrative arc came from Orange Award-winning Naomi
Alderman) then you'd have been hooked. There were plans to run the game for at
least three seasons, but after the first season took twice as long to complete
as expected, Adrian and the team decided to pack it in.
Lessons:
1. Letting the imagination run wild was
brilliant fun, but very costly e.g. the cards themselves were ridiculously
expensive (due to stock, size, smells) and the live events (with actors,
logistics, venues, props) became far too complicated.
2. All of this interactivity was possible
before web 2.0 (Facebook was still very small and Twitter didn't exist at all)
so there's no need to be slavish to these kinds of platforms. Nevertheless
crowd-sourced Wikipedia pages and Google maps mashups inevitably became
essential reference tools for creators and players alike.
3. Perplex City was very exclusive – it was very difficult to drop
in half-way but that did mean for the long-standing players the game was
incredibly satisfying, so for his next project We Tell Stories (with Penguin)
he made it much easier to pick up, which increased accessibility but made it
all a bit boring.
Hannah Donovan (@han, co-founder of This Is My
Jam and formerly led design at last.fm) is picking up the online music
space and moving it in a constrained, high quality and slow direction. One of Hanna's pet hates is the completely
boring, rapidly-spreading, rapidly-vanishing data trails on social media
websites populated with "X is listening to Y". For her, this is an example of the social
web's tendency to spread culture rather than making it. Hannah got some nostalgic laughs by bringing
up a screenshot of Lily Allen's MySpace page from 2006 and using it as an
example of an individual personalised creation. Instagram has something of that
too, with it's special one-at-a-time experienced.
So This Is My Jam is about sharing your favourite song of the moment and
listening the same great music hand-picked by people whose music taste you
respect.
Lessons:
1. Constraints are important - one song at a time was an important and
ethos-defining decision. Don't think of designing in constraints as a negative
thing. Scarcity can breed...
2. Quality beats quantity every time. Rather than thinking that you need
to be sharing in real-time the newest new thing, a lot of Jam users are posting
old classic tunes. There is a lot of new content, but that doesn't mean it's
better than the old stuff.
3. Slowness is a rare but brilliant thing. Rather than constantly-updated
twitter streams, Jam is all about giving people time to enjoy music. The music
player doesn't have a seek function ("songs are only 2 mins 30secs on
average so do your friend the favour of bothering to listen to the whole of
their favourite song") and you can keep your jams online for up to one
week (the average is three days).
I haven't given This Is My Jam a proper look yet – just want to get
this blog post done...
James Stewart (@jystewart, tech lead for gov.uk) has had a the unusual
experience as a technologist of becoming a civil servant, so I might have been
guilty of paying a bit less attention here. His project, which is very much
ongoing, is the creation of a single site for the whole of the UK government
currently in Beta phase at https://www.gov.uk.
With a website aiming to cater for the whole spectrum of society – from the
average joe who wants to find out how to renew their tax disc as well as the
more complex benefits claims to the active citizens who want to read the latest
cabinet minutes – James very much focused on making decisions at the top
level which were as generic as possible. Previous attempts at rationalising
government websites had been all about site architecture, whereas James and his
team had the humility to acknowledge that .gov.uk will never replace Google as
the default home page – they just needed to build trust in the whole
service.
Lessons:
1. Don't only iterate products, keep revisiting core principles.
2. It's vital (but increasingly difficult at scale) to work across
disciplines
3. Plan for what's next. Reaching major
release dates is great, but when you're managing a beast this big you can't
afford lost weeks in the "aftermath" period.
