Open Tech 2009

Recently I went to Open Tech 2009 “an informal, low cost, one-day conference on slightly different approaches to technology, democracy and community. This year’s theme is “Working on Stuff that Matters”. This year it was held at University of London Union, central London.

This isn’t your usual web conference, not £100+ ticket a head, no direct mention of code itself, no programming techniques. Just a group of people passionate about a subject.

There were 3 streams, with a couple of odd ‘extra’ sessions, so these are just the ones I attended. Check out the full schedule. These are a combination of memory, discussion and scribblings in my notebook written up either the day after - or a week and a bit later. As such the quality and accuracy may vary…

Community and Democracy in Hijacked Space

The day started with a talk a guy from Space Hijackers, a group of activists who do subversive and sometimes surreal protests. They’re battling “constant oppressive encroachment onto public spaces of institutions, corporations and urban planners”, there was a real impression of them caring deeply for the city they live in.

They talked through a couple of recent events, to better explain what they do. A Funeral for the East End was an evangelical march to a newly opened Starbucks in Whitechapel (just down the road from me). Another was May Day 2008/1708, a recreation of a proper 18th century May Day celebration, costumes, drinking, debauchery and all.

A few interesting points came up, how to stay just on the right side of the law. One example was the May Day event, which was promoted as being co-hosted by the Police, who would provide a good old fashioned crackdown. Because of this the Police were given strict instructions not to shut down the part. Some asked how these events affect regular people/businesses in the areas they happen; Apparently the last thing Space Jackers want to do is piss off local businesses or regular workers in stores they’re protesting.

Does Freedom of Information work? You Bet!

Heather Brooke has been campaigning and fighting in the courts to free MPs expenses data using the Freedom of Information act. Long before the Telegraph broke the story with leaked reports, or the Guardian’s crowd sourced analysis, she’s been fighting to bring this to light. There was a recent Guardian article that does a much better job of explaining the story, so read that if you haven’t already.

Unfortunately, most of her talk was explaining the story in that article, so not a lot was new to me. One interesting point raised was that now is a great time for freedom of information. In the wake of the MPs expenses scandals anyone seen trying to suppress information looks very shady and getting FoI requests granted will be a lot easier. Another highlight was the comparison to Washington where expenses data is entirely open and it’s not abused. Why? Because the politicians know that at any time a member of the public, or a journalist, could check the authenticity of their claims. Accountability = honesty.

Reminds me of a 1984 quote, but in reverse; “There was of course no way of knowing whether you were being watched at any given moment. How often, or on what system, the Thought Police plugged in on any individual wire was guesswork. It was even conceivable that they watched everybody all the time”

Making things happen

The titular session of the day started with Tom Loosemore from 4iP asking what are the barriers to making things happen? It used to be finding developers will to work on these projects, but now? The law. The biggest stumbling block faced by projects is legality, copyright and licensing.

A couple of interesting points were; the angrier the letter, the less likely they are to press a legal matter (if they could, they would already) and be nice to lawyers. Mostly yours, having a good lawyer on your side is invaluable, buy them sweets if you need to.

Following on was Tom Steinberg of mySociety, with a really interesting and original session. With the MPs expenses review there is a change to take proposals to the review board and give suggestions. The session started with a blank document and Tom built up an ideal-world wish list of features that would make up a public expenses system. All the suggestions came from the audience, it was a really interesting format for a session. Very in keeping with the mood of the day in general.

Next up was pretty much the same thing but for Free our bills. A few common themes between two emerged, most notably version tracking, attribution and annotation. 3 very simple ideas that, I think, need to be baked in to any system like this, it’s utterly essential. As I mentioned earlier; Accountability = honesty. These concepts are at the heart of transparency, from these everything else can flow.

10 cultures - Bill Thompson

The session title here is a geeky joke, a binary play on the Two Cultures - A lecture/thesis by C. P. Snow, something I wasn’t familiar with before. The basic idea was “the breakdown of communication between the “two cultures” of modern society — the sciences and the humanities — was a major hindrance to solving the world’s problems.” Ie., if those in politics, in power, better understood science and technology then humanity would be much better off.

Bills main point was asking how we can bridge the divide between these two cultures today? It’s pretty clear that government doesn’t “get” tech very well. This is because there are a whole generation growing up without a basing in these technologies, suggesting we need to scrap ICT in schools and replace it with Comp. Sci. - I don’t agree here. People don’t need to know the details of code, just understand what it does more clearly.

He suggests also that programming needs to be ingested in to all culture, not just geeks. That we need good role models that those from the “other” culture can identify, surprisingly RMS is not a good candidate here…

The only let down in this session was the overly jokey tone. While a presentation full of laughs is always a good thing (see Goldacre below), it felt like a lot of the audience questions were dismissed with a bit of a joke, which is a shame because some of those question were very well put. Particularly a comment suggesting that learning programming is not the best way to get in to the ‘geek’ cultural mind set - or that geeks need to be more politicised (though that’s not so much a problem with the audience here)

Beyond Bad Science - Ben Goldacre

The superstar of the afternoon, I suppose. He’s written the Guardian’s Bad Science column since 2004, which has spawned a very widely read blog and a book or two. He generally takes on quack medicine and the horrible misrepresentation of science in the media.

Science reporting has two big problems. 1) Regurgitating press releases without checking their validity and 2) sensationalism, popular stories have higher readership and thus make more money.

Next he discussed the power of blogs, why Bad Science and its ilk have had such importance. They’re instant, can focus on more detail than MSM and most importantly they can overlap and build up bigger pictures. One mildly angry blog mocking a science article is a nice read, a network of them sheds light on greater problems.

There was also a quick mention of Spine Wizards suing Simon Singh, and how this is just putting the spot light on them ever more. It’s essentially a mini Streisand effect.

He mentioned someone data-mining sites of the members of the British Chiropractic Association, looking for terms/services advertised and comparing against the code and services endorsed by BCA. Also, how these organisations have taken their sites down to try and perturb their opponents. Taking down their own sites as a defense. Moral of the story? Don’t fuck with geeks. And these are just the ones doing it because they believe in it, not for the lulz.

To be honest, i’d not really followed Bad Science seriously. I’d read the odd article I was linked to, but I wasn’t overly familiar with Ben or his style. Frankly, he’s fucking hilarious. If being a dual-class Doctor/journalist doesn’t work out I’m pretty sure he could do stand up comedy. I’ll definitely try and see more of his talks in the future.

Web of Power - what’s next after politicians?

This talk was the highlight of the day for me, taking 4 full pages of notes. Presented by Richard Pope and Rob Mckinnon, two great speakers, and both absolutely passionate about what they do. The theme of the session was Power. Who has it, and what we know about them. Slides available.

The diagram (as pictured at the top of this post) is from Who runs this place by Anthony Sampson. A wonderful book, which i’ve since bought and plan on reading very soon - I suggest you do the same. There was a nice intro showing various powerful people moving between the different spheres of power. Did you know Tony blair is currently on the board of 2 banks? I didn’t.

It also made the excellent point that “our” coverage of these areas of power is very lop sided. Sites like those from MySociety do a great job of covering government, whitehall, MPs, but we that’s only a part of it. What about business? Well at rewired state some folks made Companies Open House, an open and API-providing version of the official Companies House site, giving data about registered companies. But that’s limited, some of the data requires payment, like the board of directors for a company. There’s a lot more work to do here, and this talk was a brilliant call to action.

In comparison to the UK the USA is doing an excellent job with more transparent data. For example usaspending.gov displays data on Federal spending, including which contractors the money was spent with. Apparently the UK gov. doesn’t track company numbers for suppliers, this makes building anything on top of this data really hard. Unique identifiers (such as company number) are vastly more useful than a string of “BOEING”, which of the probably half dozen different Boeing companies did you mean?

This got me thinking about the usefulness of sites like theyworkforyou, or companies open house. They don’t just provide information, they provide persistence of both identification and destination. For example, with a companies house number you can link to the companies open house site. Incredibly simple. But is there a way to do this for, say, a member of a companies board, or a product? The usefulness here is an established unique identifier (even a defacto one) and a (at least semi) authoritative site that represents that resource.

Is Companies Open House the perfect representation of a company in the UK? Probably not, but if I link to a company there I can be sure that you’ll find it useful and could find information more easily from there than here. Now this starts sounding like the slightly mucky world of the semantic web - lowercase intended. These sites aren’t 100% percent perfect representations of resources, they don’t need to be as rigid as RDF, they are essentially cow paths. They represent an emergent mapping of the underlying power system. They give just enough semantic links and meaning to let someone scrape a combination of these sites and show wonderful data and results. The more resources we have mapped the better.

There was also a quick discussion of Richard’s site Straight Choice which tracks election leaflet claims, as so many are made but forgotten post-election. It could also be used in the future for fact-checking those leaflets, validating the claims and exposing parties making fraudulent ones.

Another interesting point was super market loyalty cards. I don’t have one, I find them annoying and invasive. It’s a lot like scrobbling on Last.fm, you tell the super market exactly what you buy and when you buy it. We use the term attention data, and this is exactly the same concept. This gives them the power to do massive market analysis, it’s very very powerful data. However, it’s of no use to the user. Sure they might get some vouchers to save a few quid once a year, but I want access to the data itself, that’s way more useful.

And to tie in to my previous points about identification and authoritative linking, it should alert me to news about companies I buy from most. Or companies against my political and moral positions. Is Company A, who I spend £X a month, actually evil (subject to personal preference). Are they killing children in 3rd world countries, should I boycott them? Or a less political example, I buy Product B 20 times in 3 months, maybe I should consider buying it in bulk? Or, health analysis based on what you’ve bought, which incidentally is something Last.fm’s healthcare provider do. You earn healthcare points if you give them your club card and spend X amount on fruit and veg.

Getting the data is hard though, it was mentioned some show you data through their own sites, or that you can request exports, though the format and availability is unclear. This is something that really needs looking in to. There’s an absolutely killer oppertunity for a 3rd party site to do this amazingly well, assuming you can get the data in. I hate to say it, but in this economy a site that offered money saving advice based on your loyalty card purchase history, something the supermarket itself would never do, would be extremely popular.

A recent project of mine has been taking an export of my bank transactions and trying to build a useful, personal, site around them. Online banking sites are notoriously bad at being useful - usually under the banner of security. What I want is a good site or app based on my transaction data. There’s incredibly interesting and useful data in there. I can work out my fixed outgoings, my balance at the end of each month, if i’m trending up or down, how am I saving, my big purchases, my frequent purchases, travel costs. And that was just with the quick and dirty CSV/PHP version.

What I really want to do with that data is get it in SQLite, wrap a really power framework around it and turn it in to a proper account management app that I have 100% control over. This is definitely going to be one of my main project for the next few months, so expect more posts about it. The same principle applies to this super market data, i’d happily use a card if I could get hold of the data easily. Basically, just let me get my data (easily) and do awesome stuff with it.

Ephemerality? Real time web vs persistence

My second favourite talk of the day. Lots of cross over with ideas i’ve been thinking about recently and projects I’ve been working on. The slides also featured a PKD quote (slightly re-appropriated), which is always good.

The opening discussion was “do we need to remember everything?”. Forgetting is useful, as humans we rely on it, we forget details and remember the important bits and a general overview. Do I really care what track I listened to 3 months ago? 99% of the time, no I don’t. I care far more about the abstract representation, what kind of music was I listening to 3 months ago. The further back you go the less you need to care about detail, 3 months ago I’m interested in what artists I liked, 3 years ago genre’s are more useful reference points.

This abstraction gives context for the time period, “The summer I went to SXSW”, “The year I discovered hip hop”, etc. It’s really quite powerful. There was mention of a version of a complete works of Shakespear which includes notes detailing the nuance of the language of the time. This gives you a much clearer context to understand the work. But what if you apply this idea to not just you, but your friends too, then you get social context, or a “social atlas”.

This is something that can operate on smaller time scales, I really want to know the abstracted trends between my friends for the last week, or month. It’s something we fail to do well at Last.fm, personal charts are great, friends charts are much harder but potentially incredibly useful. Have half a dozen of my friends said they’re attending the same gig? There’s a very strong chance I’d want to go to. The same applies if lots of friends have discovered the same artist recently.

However, sometimes you do care about a very particular date or detail, not the abstraction. This can be hard, humans don’t remember all the details, we remember bits. We remember a detail, that reminds us of another detail and bang, we remember the information we wanted.

I’ve found myself doing this a lot. A really interesting example was looking at my bank statement “Holy crap, when did I take £80 out of a cash machine at 10am?”. I had no idea, fraud perhaps? I checked through iCalendar, nothing. Most of my calendars are feeds for future events from sites like Last.fm, Facebook, Upcoming, so there were pointless and going through each site looking at my past events felt like hard work. Instead I checked twitter, went back through my history and discovered a tweet from that day - visiting Portsmouth for a day trip. This immediately triggered the memory of taking out a bunch of cash at the start of the day in case I couldn’t find a cash point, problem solved.

There’s a lot of cross over here with my lifestream app. The current version is fed by RSS and is as transient as it’s source. Ideally it should record everything locally, that’s something I really need to do. I’d also like to do the same and permanently log what my friends have been doing too. It’s very similar to the idea behind friend feed, but ultimately more idealistic and personal.

Location and Privacy

Next up was Gary Gale, from the Y! Geo technologies team. He discussed the changing attitude towards privacy online. Online privacy is a relatively new model, and compared to existing offline privacy models there’s an increasing move towards opt-in than opt-out.

For example, in the UK if you’re on the electoral register you have to actively opt-out of being on a commercial address database. Equally your telephone number is available publicly unless you go ex-directory. This is pretty normal here, no one’s too bothered, which is interesting considering both these sources of data can be used for very physical annoyance, post-spam and direct dialing calls.

Compare this with more recent online examples like Facebook, Dopplr or Fire Eagle. They tend towards being more opt-in, with granular options for who can see what about you. I’m not sure I entirely agree, while opt-in makes a lot more sense for geo-aware applications, the majority of sites still tend towards being totally open, with options to hide your information.

An interesting point about geo-privacy was that you need to be able to disable it. That’s great, but if it’s something you use 99% of the time then there’s the risk of guilt by omission.

Local web beyond the hype

I only caught the end of this talk, after leaving the energy talk in the main room. There was some nice discussion of different community/local sites. There was a great example of a small village with a simple wordpress blog that had ~75% readership. I’d like to see the full set of slides for this talk if anyone has them?

No2ID and Open Rights Group: Intercept Modernisation Programme

This session started with a “very special guest”, Sir Bonar Neville-Kingdom a senior advisor in Whitehall and Data Sharing Czar. A rather prim and proper gentleman in a full suit took to the stage and started giving talking, at one point quoting some ancient Sumerian(?) poetry. It took a minute, but it dawned on the audience he was a pretty elaborately staged troll, put on by the Open Right Group.

The whole session was about IMP, the talks were entirely useful, covering the idea, why it will fail, etc. To be honest, by this point in the day, in a very hot room I kind of zoned out. The talks just didn’t strike a chord with me, there was little interesting content. Showing the flaws in the proposed system is best left for campaigning, i’d have preferred a discussion on how much we need a system like this and how we’d design a better one.