Dragon Trainer begins

Good news: a research project I helped to write has been approved for funding by the European Commission’s Future and Emerging Technologies program. The project is led by one of the scientists I admire the most, David Lane, and rests firmly in the complexity science tradition associated to the Santa Fe Institute. We intend to attack a big, fundamental problem: innovation is out of control. Humans invent to solve problems, but they end up creating new and scary ones. Which they tackle by innovating more, and the cycle repeats itself. Cars improve mobility, but they come with global warming and the urban sprawl. Hi tech agriculture mitigates food scarcity, but it also gives rise to the obesity epidemics. To quote one of our working documents:

While newly invented artifacts are designed, innovation as a process is emergent. It happens in the context of ongoing interaction between agents that attribute new meanings to existing things and highlight new needs to be satisfied by new things. This process displays a positive feedback […] and is clearly not controlled by any one agent or restricted set of agents. As a consequence, the history of innovation is ripe with stories of completely unexpected turns. Some of these turns are toxic for humanity: phenomena like global warming or the obesity epidemics can be directly traced back to innovative activities. We try to address these phenomena by innovation, but we can’t control for more unintended consequences, perhaps even more lethal, stemming from this new innovation.

We want (1) build a solid theory that concatenates design end emergence in innovation and (2) use it to forge tools that the civil society can use to prevent the nefarious consequences of technical change. It does not get any bigger! And in fact we got a stellar evaluation: 4.5 out of 5 for technical and scientific excellence and 5 out of 5 for social impact.

The project commits to building Dragon Trainer, an online community management augmentation software. The idea is to make a science of the art of “training” online communities to do useful things (like policy evaluation), just as you would train an animal too large and strong to push around. I am responsible for producing Dragon Trainer, and it is quite a responsibility.

I am superhappy, but worried too. Taxpayers foot most of the bill, and this makes it even more imperative to produce the absolutely best result we can. I will need to work very, very hard. I am seriously thinking of devoting myself to full time research for a couple of years starting in 2012. Does this make sense? What do yo think?

17 thoughts on “Dragon Trainer begins

  1. Pedro Prieto Martín

    Alberto, the first: congratulations!! I hope the Dragon really behaves and learns to do magic.
    The second I would like to point you a potential fruitful collaboration, for the future. As you know, the Asociación Ciudades Kyosei is starting the creation of Kyopol (“Symbiotic City”), a system that aims to reinforce civic engagement and citizen activation. We are creating it bottom-up, with the help of representatives of all future users of the system (citizens, social movements, associations, local media, public officials, elected representatives, politicians…) and we are considering it not just as a technical device (ie: functionalities), but also taking very seriously its “procedures” of use, and an institutional framework to improve its governance. Much of this is very aligned with your approach to “train dragons”.
    Well, it is to be seen if it will attract a “critical mass” and thus the utility it generates grows enough as to atract the normal citizens. But… we hope it will. We are going to pilot it in Spain, were the 15M movement (the “indignados” or the “spanishrevolution”) are in a huge need of it). They could act as the innovative “critical mass” that later attracts an important portion of the citizenry, which in turn will attract media attention and government involvement.
    The system aims to help in the “policy development”, but from a more traditional perspective: we do not aim to create robots that suggest targets and explanations, but make extremely easy for governmental and non-governmental to poll different sub-sets of society, to perform easy or complex “opinion mining”. This could prove to be a very valuable approach to design, fine-tune and evaluate services and policy making.
    Well, IF our system works… it could be a very good environment to put your “dragon” to work. Combine the “cyber-politician” we want to create (by enhancing the senses and capacities of human actors), with your “AI robot-friend” that provides network analysis and tips (if I understood well what your “robot” is supposed to do 🙂 ).
    We have still a lot of work to do. But… well, maybe in the future we can really think about synergies and collaborations.
    And again once more: congratulations.

  2. Enrico Ferro

    Che bel progetto! Complimenti. Seguirò appassionatamente i suoi sviluppi. Avete già messo in piedi un sito web? Ti sarei grato se mi potessi tenere aggiornato sui suoi sviluppi.
    Grazie a ancora congratulazioni!.

  3. Federico Bo

    Dedicatici, dedicatici 😉 Ti seguo con attenzione anche per la nascente community di Cineama (che sto facendo “ammaestrare” seguendo gli insegnamenti della scuola (università) Kublai…)

  4. Tiziana

    Sono felice che “Dragon Trainer” sia ora un progetto reale! Anche la fantasy può essere d’aiuto, alle volte. Studiare o produrre? Raccontare o agire? Un dilemma che non si risolve con facilità. Fare l’uno e l’altro a volte è impossibile. Bene allora fermarsi ogni tanto e immergersi in una cosa alla volta. Se questo è il tempo dello studio, va bene così! Auguri!

  5. R. J. Hucker

    the picture of the dragon in the beginning is the symbol of Backtrack, a Linux distro…it is a little misleading, it should not be bad to change it.

  6. stefano maffei

    anche io sono interessato a seguire l’avanzamento del progetto.
    maila quando partirà il sito…e considerami come un possibile “uditore” nelle parti pubbliche…mi raccomando…
    il tema laneiano e tuo dello spazio dell’innovazione agente/artefatto/sistema lo sento particolarmente per formazione e applicazione disciplinare (fare design è…in una prospettiva teorica…fare questo)
    in bocca al lupo..

  7. Paolo

    congrats and few suggestions 😉

    1) the most important is “KISS”! 😉 http://en.wikipedia.org/wiki/KISS_principle

    2) the most difficult part will be “adoption” of your platform (reaching the critical mass). My suggestion is: KISS and devote a lot of attention, especially in the beginning, to incentives: why should your users use the platform? What will they get out of it in the beginning?

    2.1) Eat your own dog food: be the first user of your products! If you don’t use a platform you build (because you don’t find it useful), it is very unlikely other people will use it! Be the first users of your platform! This means: for coordinating the EU project, use the platform you are creating itself! Even when it is just a simple 0.1 beta! And remember, it should remain Simple! 😉

    3) making the platform open-source will help a lot (for your image, for adoption, for contribution)! and will justify a bit more taxpayers footing most of the bill 😉

    4) do you already have mockups of the platform? I would be interested to see them.

    5) I’ve been recently researching the Wikipedia community, trying to apply social network analysis too (I saw that you refer to Wikipedia as an interesting case in your other Dragon Trainer post). Not easy but surely the most interesting case you can study large-scale and empirically: every edit by every user to every page in the past 10 years is available and the same platform, mediawiki, is used by 280+ different communities, by language, so you can check if your results hold cross-culturally, cross-linguistically or at least cross-different-communities. Moreover users both create content (main activity) and also create-update their rules (pages in the Wikipedia namespace), moreover there are pages in talk namespace and user talk namespace which can be easily analyzed from a social network perspective.

    With this regard, there have been recently some papers by gleave lento welser smith trying to automatically assign “social roles” to users of social networking sites (and Wikipedia!) based on their contributions and social networks. I think their goal is to solve what you write “Community managers, myself included, are trapped in this dilemma: practically the only way we have to figure out the social dynamics in our communities is to spend an unreasonable amount of time participating in them, and we try to steer them by rhetoric and persuasion. We end up navigating pretty much by gut feelings.”. One of the authors is in Facebook I think and you can surely imagine that it is hard, for people working at Facebook, trying to understand the entire Facebook community and how to keep it healthy. I’ve spoken with people at Wikimedia Foundation and this is surely their main concern as well: they sit on top of a vocal, creative and powerful community of people (“sit on top” is not a great metaphore, figuratively speaking) but they are not able to understand what these millions of people like and what they don’t like, what attracts and keeps them in the community and what makes them leave.
    So again, Wikipedia would be a great example to study 😉

  8. Ivan Vaghi

    Paolo, one of the initial challenges is to understand the basic raw underlying model that we are going to analyze. As you point out a social graph can be extracted from Wikipedia, although it is an implied model, not an explicit model. What is in your experience an effective pragmatic ontology to build upon? I was thinking Actors + Relationships, with decorating attributes, but I was also considering adding Artefacts, as boundary objects mediating communication and interest between people. Do you think it is worth keeping track of Artefacts as mediators and vehicles of communication or should we throw them away once we export/scrape the data and keep only the relationship?

    1. paolo

      I would suggest to keep it extremely simple! Actors + relationships is enough (it is already very complicated to extract metrics out of them) and usually data coming from online communities are already too noisy at this level 😉

      Weighted and directed: of course take into account that relationships must be modeled with their weight (how many messages did we exchanged? what is my level of trust in you, from 0 to 10?) and with the direction (I appreciate Steve Jobs as 10 out of 10 but Steve Jobs does not (did not) express its appreciation about me (because he does (did) not know I existed 😉

      Longitudinal: Analyzing the community (nodes and edges) as they change over time is great and needed too! It is also a research challenge still very open (while static networks are already extensively studied)

      Explicit vs implicit: not always explicit relationships are the best (sometimes they are not “realistic” and relying on the real activity can be more close to reality). About “realism” of expressed relationships, you might want to skim over my “A Survey of Trust Use and Modeling in Current Real Systems”, a survey of Epinions, Ebay, PageRank, P2P systems and how they express trust relationships

      A dataset I released that is weighted, directed, longitudinal and explicit is a network I extracted from Advogato daily about how much software developers appreciate each others (but Advogato is not used too much currently, certainly less than Facebook 😉
      I haven’t played with it too much yet. You can find it at http://www.trustlet.org/wiki/Advogato_dataset

      1. Alberto Post author

        That’s very generous, Paolo! Thank you!

        To your list, I would add: multiplex networks. The same nodes (people) can be connected by several types of links (comments, friendships, participation to the same groups etc.) giving rise to “layers” of connectivity. There is information to be milked from the correlation between the different layers!

        Say hello to Napo 🙂

  9. Ivan Vaghi

    Engaging the community is one important point that we are pushing, but I can see a trade off here between allowing the researchers on the project to be effective asap, and making it simple for general adoption. Definitely something to think about.

    Paolo, it would be great if you could point out some people you think might be interested in our work (and maybe getting to use/contribute code) once the project gets started.

  10. greenguy.wallinside.com

    t figure our way out of our consumerist addictions we will eventually have to go cold turkey.
    This can impact both worldwide and local patterns of climate
    and atmosphere-ocean circulation. While the release
    of greenhouse gases may not cause immediate illness, the sum total of what has been
    released into the atmosphere is starting to cause problems around the globe.

  11. cavs

    Would love to know how far your are into the dragon trainer project. It sounded like one massive idea. wish you succeed in it.and when you do am sure you will be back here and share with your blog followers like me.


Leave a Reply

Your email address will not be published. Required fields are marked *


This site uses Akismet to reduce spam. Learn how your comment data is processed.