This weekend, I spent the day at OpenTech 2010, with several friends, enjoying a day of interesting and thought-provoking geek talks (only slightly marred by a re-acquaintance with the concept of Saturday mornings and by my stomach being awkward). My notes from the talks that I actually took notes in are pretty sketchy and lacking in detail, but I figured I'd post them for anyone who's interested. I'm gonna list all the sessions I went to, even where I didn't take notes. Obviously, I'm typing these up well after the event, so any errors in the interpretation of my notes are entirely mine and not due to the speakers.
It's worth pointing out that the OpenTech schedule page now has links to audio for most sessions and notes and slides for several others; I've linked to slideshows where appropriate.
data.gov.uk — process and properties
As they describe themselves:
Transparency is at the heart of this Government. Data.gov.uk is home to national & local data for free re-use.
All their work is done with open standards; where there wasn't an open standard, they've created a generic one, then specialised it to their needs. They have URL-based schemas for everything, based on official IDs, which lead me to worry about discoverability; as this was the first session of the day, though, I completely didn't have the higher brain functions adequate to ask any questions on the matter ;o)
They use a package called Freebase Gridworks to manage all this data, and they've deliberately made it relatively simple for civil servants to import spreadsheets into it, generating JSON, which can then be copy/pasted into Gridworks, which can then be exported to RDF and other formats. They mentioned TSO's Data Enrichment Service, which looked particularly interesting, but I can't find it on a quick google. This was URL searchable — adding, for example,
?religiousCharacter.label=Roman+Catholic to a search of schools filters the list to include only RC schools.
It's easy to create bespoke APIs around the data, such as the Simple Spend Explorer (which I can't find on Google). Dan (?) built a site that graphs the relationships between API data (such as the org-chart at BIS, with a dynamic Ajax and JSON UI.
legislation.gov.uk — built on the API
Legislation.gov.uk is one of my favourite public-data sites of late. So much nicer to use and a much more pleasant UI than its predecessor Statute law.gov.uk, it is based on a RESTful "Linked Data API", where URLs are eminently guessable and hackable and machine data can easily be fetched by varying the URL:
1929/25/9/contents is the human-readable contents of Magna Carta 1297
1929/25/9/contents/data.htm is the same information in HTML again
1929/25/9/contents/data.xml is the same information in Atom XML
1929/25/9/contents/data.rdf is in RDF
1929/25/9/contents/data.xht is an XHTML snippet (without the
bodyaround it and so on)
1929/25/9/contents/data.pdf is an Adobe PDF
Can Twitter Save Bletchley Park?
Dr Sue Black, from the University of Westminster School of Electronics and Computer Science (@Dr_Black) is a total fangirl for Bletchley Park and has been instrumental in attracting attention to the plight of one of the most important sites in the history of computing and of the British war effort.
Because I already know a fair amount about the issue, I took almost no notes at all. Dr Black did mention, though, that she'll be putting her presentation online and her site and blog have copious links. The one note I did make, though, was to look up the paper she presented to the Museums and the Web conference in April: Can Twitter Save Bletchley Park?.
Feedback for Emergency Doctors
(Slideshow) Dr Carl Reynolds firstly ran us through some medical history:
- In 1747, James Lind ran the first clinical trial, proving his theory that citrus fruit cured scurvy.
- In 1847, Ignaz Semmelweis discovered that handwashing (with chlorinated lime) decreased mortality from childbed fever in new mothers.
- In 1854, John Snow, one of my personal heroes, effectively founded epidemiology, discovering that the outbreak of cholera in London was transmitted by the fæcal–oral route, persuading the parish commissioners to remove the handle from the contaminated water pump.
- Currently, around 250,000 people die in the UK every year from nosocomial infections.
Part of the problem is, apparently, that the NHS has little capacity to learn. There is no routine feedback — unless the patient returns (not usually the case in emergency medicine), doctors never find out if they made the right diagnosis, unless the patient complains, sues or dies. It's imperative to focus on changing systems to resolve problems — blaming people means that failures won't be reported, which prevents the NHS from learning from its mistakes.
Dr Reynolds runs a project at the Royal Free that seeks to change this by making sure that doctors get weekly feedback on their cases. But to do this was a struggle. Why?
- Learned helplessness
- Lack of support for innovation
- The need for the personal computer revolution to hit healthcare — that medical staff are disallowed from using their own devices and required to use dated NHS hardware means it's difficult to do any UI well.
Gavin Starks: Climate Change update
Bill Thompson: Giving the Enlightenment another Five Hundred years
Our lunch took forever, so I missed Gavin's talk and I was too busy absorbing the awesome infoporn to remember to take notes during Bill's. Though he said we should digitise everything and not forget the metadata, if that helps ;o)
For the win: game-space and public engagement
This was, with the possible exception of Bill, my favourite session of the day. Chaired by Paula le Dieu, director of digital at the BFI, the panel included Tom Chatfield, author of Fun Inc, Alice Taylor, Commissioning editor for education at Channel 4, and, as a surprise extra guest, her husband Cory Doctorow, author, blogger and activist (like you don't know who Cory is if you've bothered reading this far). My notes here aren't comprehensive, but not as sketchy as for others.
Gold farming is increasingly prevalent; Tom mentioned the trade in virtual currency is potentially worth $100 billion by 2020 and Cory mentioned that around 400,000 people earn some or all of their living from gold-farming. Alice mentioned Entropia Universe and Raph Koster's book A Theory of Fun, but I didn't note down why (only that I wanted to look them up).
Alice spent a good while discussing how Channel 4 find that games are very useful means of reaching 14–19-year-olds with educational messages; copious examples can be found on the Channel 4 Education blog, such as Privates, a "direct, accessible and daring approach to sex education", "aimed squarely at post-pubescent boys, allowing them to explore sexual health and get better informed" — a "downloadable PC game set in various human orifices, where you battle chlamydia, gonorrhoea, Aids and more along the way". Channel 4 spend over half of their teen education budget on education-themed games development. Later it was mentioned that, in many cases, public-service broadcasters should be spending more of their budgets on games-development and internet projects than on television, given that that is how to reach many target audiences (particularly the 14–19yo teen demographic that is Alice's educational focus).
Cory made an exceptional point (which became a top tweet) that this virtual economy should not be overlooked and looked down upon — "virtual wealth" (ie in-game currency) is not fundamentally different from collateralised debt obligation derivatives, which are, after all, abstract fictional money in most senses.
I was bemused at the evident disdain for Facebook gaming that seemed to fall between the lines given this next point: the World of Warcraft EULA (see clause 6) and terms (clause 17) are a perfect example of authoritarian abuse. Blizzard require all users to install and run spyware, to prevent from cheating (and from bot-derived virtual wealth) — in order for the server to be a fair referee, it needs to be able to trust the endpoints as being tamper-resistant.
Gaming has become more and more about limbic, rather than hedonic entertainment; we are increasingly craving to continue playing. Cory made the point that democratic games design wouldn't work — games often depend on repetitive tasks known as grinding in order to advance skill levels; noone would vote to increase the amount of grinding in a game.
- Is public data too boring for games?
- Alice: Not at all, it's just how to sell it is difficult. Bear in mind the amount of information a player picks up in order to play something like WoW to level 80. Understanding politics and democracy are nothing compared to that, you just need to set up the learning properly as a part of the game mechanic.
- Cory: The University of British Columbia did a study that showed the average Canadian child could recognise 400 Pokémon, but only 4 birds. So Dave Ng at UBC put together a game called Phylomon, a trading card game with the same mechanic as Pokémon that rewards players for learning about species. He mentioned a first-person shooter that Alice developed where the game-map is made up of public museums.
- Tom mentioned Classical Greek memory palaces — human psychology is such that spacial information is more easily retained.
- (Question I didn't record)
- Alice: Around 80% of all games are unfinished by players. For channel 4 they have games ranging between 6 million players and 80 million plays for 1066, a military strategy game, to Smokescreen with around 150,000 players, that teaches about the privacy implications of social networking as you stalk someone on "Fakebook" and "Tweeter".
- Tom: One of the weaknesses of interactive media is the lack of serendipity, so it's easy to end up preaching to the converted. Because gaming and fun are universal, this makes it easier to avoid this, and increase the sense of serendipity.
- Cory: That said, my media intake is increasingly diverse and serendipitous over the last 15 years, and only partly because I edit BoingBoing
- Tom Steinberg: Does anyone know of a game that produces and actual out-of-game side-benefit, where the externalities are something useful, in the way that Recaptcha both helps reduce spam but also helps with OCR scanned texts?
- Tom: Fold.it, the protein-folding game, has direct medical benefits
- Cory: Google Image Labeler, which helps Google develop its image recognition algorithms by working on the principle that if two strangers agree on something, it's probably gonna be accurate.
- (Question I didn't record)
- Someone mentioned Jane McGonigal, Jesse Schell and his talks about the "gamification" of life (TED, CBC)
- Alice: The formula at places like Zynga is around 2 data analysts for every 5 production staff (developers, project managers); it's the triumph of the A/B test.
- How can we use games to turn people into social contributors, rather than just consumers.
- Tom: A lot of games can be paths to action.
- Alice: At Channel 4 we are soon to announce a fashion-related game, encouraging teens to swap, recycle and exchange clothes, rather than going to Primark. To an exent, you just need to design social contribution into the game mechanic.
- Cory: It reminds me of one of Alice's favourite aphorisms: "Any number on the wall goes up" — if you give people points for doing something, they'll do more of it. So the question is: how do we create leaderboards that reward people who innovate in-game and encourage lateral thinking?