Last weekend I went to the Barcamp Leipzig. Location, weather and the attending people were brilliant, so it was a weekend well spent. Barcamps are also called "unconferences", as they are conferences without a plan. There is no predefined schedule, instead, anybody from the audience can declare to give a talk or host a session on any topic they fancy. Most sessions are about computer stuff, but it is not mandatory. For example, at the barcamp munich there were also sessions on magic tricks and night photography. Many people decide to host sessions without having prepared very much, but it usually works well, because it ends up more being a discussion with the attendees of the session, rather than a frontal lecture.

There is always something that puzzles me about events like that, which is the way I, and maybe most others, socialize. Why do I talk with some people, but not with others? Usually in the beginning I talk to a lot of people, and nearing the end I hardly approach anybody new. It wasn't that extreme this time, but there were still some people that I have never spoken to over the whole conference. Last year I went to a one week workshop in Amsterdam, and there were people I never talked to once. What I wonder: are the people I talk to the right ones, and if so, how did I know? Not to be misunderstood, I didn't meet anybody I didn't like at the barcamp, but there might be people that have the same interests and would really click with me, becoming co-founders for the next startup, and I might not even have talked to them. In any case, if anybody reads this: I am happy to hear from you, even if we didn't exchange contact information.

This question gave me an idea for another project: analyze social graphs available on the internet to see how people socialized at barcamps. For example, there is a twitter account for the barcamp leipzig, and it seems most attendees who also use twitter decided to follow it. Now I could look a these people's friends lists on twitter before and after the barcamp. If they became followers of each other shortly after the barcamp, they probably met at the barcamp. Then I could analyze their tweeds to determine their interests and similarities (maybe some other social network would be more suitable, for example Xing lets people list more specific interests than twitter).

Finally, a brief overview of the sessions I went to see:

Genetic Algorithms

Kennon decided to use slides from his lecture notes to give an overview of genetic algorithms. There was nothing new for me, but it was nice to meet somebody else who has the same interests. Also, the barcamp organization inspired a good area of application for the GAs: it would be nice to arrange the session slots in such a way that as few people as possible miss sessions they would have liked to attend. Maybe one of these days I'll try to implement that for fun.

Open Street Map

I have known about open street map before, and I have high hopes for it: it is the attempt to generate maps from GPS data that is submitted by volunteers. Similar to wikipedia, everybody is allowed to edit the maps, but some geeky equipment helps. Somehow I had never given it a closer look, but the enthusiasm of the two presenters Claudius and Nathanael was infectious. Clearly, they were having a lot of fun scouting for uncharted terrain to close the gaps in the open map. For example, if they were driving somewhere and see an uncharted road, they would make a brief detour to collect the data for OSM. To me, it sounds as good a motivator as any other to go outside. Also, as I descibed in my article about GeoTags, I have become interested in maps recently, and having freely available data would be great.

I hope to contribute to Open Street Map in the future, unfortunately at the moment I don't have a capable device. Actually walking routes with a GPD logger is not the only way to contribute, though. For example, Open Street Map apparently was the first online map to provide maps of Bagdad, which had been transferred from satellite images by volunteers.
It was great to see how quickly OSM is growing - you can watch the expansion of charted terrain over time.


This talk ended up being more of a brainstorming session than a lecture. Scout24, the employer of the speaker (Cindy Beer), is apparently wondering how to ensure ongoing innovation in their company. One strategy could be to look at ongoing "megatrends" and think about what needs they could produce in the future. Examples for megatrends being mentioned were the aging of society and the increasing use of mobile internet. For me, there was no "new megatrend" that I hadn't heard of before. It would have been interesting to hear what scout24 thinks are the important trends for them, but the presentation did not get to that point because of the discussions with the audience.

Another problem is how to manage the innovation. For example, one approach would be to create a dedicated "innovation team" within the company. This could result in all the other teams becoming "uninnovation teams", though. They might decidedly not innovate, or even worse, if any other team would innovate, the innovation team could feel threatened and try to undermine the other innovating teams.
Incentives are another problem. At least I feel that I would not like to give my best ideas to a company that I was employed for. If the company would make millions from my idea and I would get nothing but my paltry salary, I would feel cheated. So some kind of reward scheme seems necessary. It is not always obvious how to quantify the value of an idea, though. Not every idea translates directly to revenue for the company. If a team in the company is responsible for evaluating the value of an idea, the "innovation team" problem kicks in again. What if you have a great idea, but for some reason, the evaluation team doesn't like it? O2 had a system like that - I think it was still better than having no means for employees to give input on new ideas at all.

For me that discussion reminded me about "prediction markets", which are apparently successfully used by some companies. The idea is to create a market similar to the stock market, but people can bid on ideas with virtual money. That way, the evaluation of the idea happens through crowd wisdom, instead of single individuals. A public example is Yahoo! tech buzz, but some companies also use such tools internally.

Another point that was raised that to look for new ideas, one should not only think about problems that look for solutions. Sometimes just going for something fun works, too. I am not sure I am convinced: making something fun also solves a problem, the boredom or loneliness of people. But maybe only looking at everything as a problem really is cropping too many thoughts.

Twitter for Corporations

Twitter continues to boggle the minds of everyone, even the people who use it on a regular basis. Since I am a fairly new Twitter user, I was hoping to gain some new insights by this session on corporate twitter. The plan for the session was to look at several examples of corporations using twitter, but because of the unreliable internet connection, it did not work so well. Instead, there was a lot of interesting discussion with the audience. Some examples for corporate twitter blogs were available, though.

Fail: the Deutsche Bahn is twittering about problems with their trains in remote locations, which can't possibly be of interest to a wider audience. However, some people in the audience questioned the authenticity of the twitter, they thought it was just somebody else posing as the Deutsche Bahn. It certainly seems to be becoming a problem, people are already registering other people's domain names as twitter accounts to mess with their reputation.

Fail: Preisbock - they are an example of twitter spamming. Whenever somebody decides to "follow" you on Twitter, you get an email that informs you so. It is not required to follow back, but generally considered the polite thing to do. So some companies simply try to follow everyone, and a lot of them follow back. It is very comparable to "trackback spamming" for blogs. However, if it fails, you leave the target of your spam with resentments against your company. Not good.

Success: I don't know, the presenters liked the twitters of bild and some other magazine whose name I forgot. Bild is of course the specialist for catchy headlines, so it is not a surprise that they do well on twitter (the main attribute of twitter being that messages are limited to 140 characters). They do the same they do with their headlines, short, catchy sentences that are almost impossible to ignore, even if you actually despise the newspaper. The other magazine was giving their writers turns to twitter about the things happening at the office. Some people liked it, but others thought it made a very unprofessional impression.

Success: Till from was in the audience (and they provided excellent coffee for the barcamp). They have a twitter account for sonntagmorgen and as an example, track people twittering about coffee and follow them. That approach seems more acceptable than the random approach taken by Preisbock, although I guess technically it is still spamming. Sadly I forgot what else they do and how successful they are. I think Till is a very good marketeer, though, so I think they must be doing well.

Success: I don't understand what betacamper does, but they were successful with viral marketing on twitter. The trick is simple: after signup, redirect the user to twitter with a prefilled form field saying "I just signed up to beta camper". Users are just the press of a button away from spamming their friends with this message, and apparently many did so. Much more effective than the old "recommend to a friend" email forms. On twitter, you don't need to enter your friends email addresses, and because of the short sentences, you don't need to wreck your brain for something interesting to write to your friends. I am definitely planning to copy this approach.

Other interesting points in the discussion: some people thought that Twitter is just the same thing as an IRC chat. Andreas told me about a friend who some years ago really used IRC the same way people use Twitter today: hanging around in one channel all the time, messaging about things like "cooking something for dinner now" or "going to sleep now". The difference with twitter is maybe that you are not constricted to a specific channel, and of course it is much more accessible than IRC. Everybody builds their own channel on twitter - most people said that twitter is more like background noise than a dicussion you follow closely. I asked some twitterers who follow more than 100 people how they do it, and to my surprise they admitted that they use filters to only see important messages. So twitter becomes a farce in the end: it is a symbolic act to follow somebody, but people do not really read each others messages in the end.

Still, I think Twitter has a lot of potential. The service itself is maybe too basic, but a lot of other pages build on Twitter, and some killer apps could emerge. I like the idea of mining the network, as I described above (who made friends at the barcamp), there are filters to see only important messages, and so on.

(to be continued - I have to leave for work now...)