Open Innovation is a catchy phrase, but I don’t think we are that close to it, as many people claim. Innocentive, InnovationXchange or NineSigma operate in the very small market, and this market does not seem to grow as fast as we would wish. Innocentive posted some statistics as of 2nd of June, 2009, so given these numbers and amount of open challenges, it’s safe to assume that as of today, around total of 1000 challenges were posted and ca. half of them were awarded. If you compare that numbers with almost 200 0000 patents issued only by US Patent Office in 2006, it gives a clear picture of the size of the market open innovation crowdsourcing companies (edit: as Jean-Claude points out in the FriendFeed comment, Innocentive and the other two companies mentioned earlier are rather crowdsourcing, not “open innovation” companies) are operating in. There are plenty of reasons why OI did not yet become mainstream (too many to list) and for that to happen, there are two important steps that we need to make first.
Open Science must become mainstream
I’ve been advocating Open Science for some time and I’m following Open Science luminaries for much, much longer. At some point it hit me that Open Science in its fullest form is not an issue that scientists can truly solve by themselves. Open Science crosses domain of Science – it’s an issue for Science, Politics and Business. We should experiment with various ways the research is done, collaborate openly, attempt to invent new business models to fund science and spread “open” meme as much we can. However, the real deal will be made between people in power from these three domains. Why this is necessary to achieve that before we may fully innovate in the open? Because in this step we will sort out all the problems we have today with intellectual property and technology transfer (both being not efficient enough for today’s standards). I cannot envision that happening in other domain – we are paid to collaborate and test ideas. This community is able to hit every major obstacle to “open” in a very short time. And once we have these obstacles removed there’s a next step:
Working models of Open Science should be tested outside of Science
In other words I postulate that whatever solutions work in domain of Science, these should be tested outside of it, in other domains. Not vice versa. Principles of Open Source software did not prove to be useful in open drug development (see Joerg’s post on the topic). Crowdsourcing will not advance quantum physics. Not all aspects of collective intelligence are working in Science. We simply need to invent working solutions within the domain first, and then test them in other domains, such as art or engineering. This step will provide another set of protocols, changes and adjustments that will allow seekers and solvers (to use Innocentive’s nomenclature) to work efficiently together crossing every domain.
Open Innovation is not a single step
I may be proved wrong by some genius that will solve Open Innovation proovedissues in a single brilliant step, but so far I believe that we need more than one to achieve this goal. And it is important to recognize that Open Science is a great opportunity to come closer to it. The sooner we realize it, the better.
![Reblog this post [with Zemanta]](https://i0.wp.com/img.zemanta.com/reblog_e.png)
Freelancing science 
Jen McCabe
July 3, 2009 at 19:13
Pawel –
Stumbled across “Freelancing Science” after this post was shared in the GenomeWeb e-newsletter.
Great discussion of open source’s failure to impact health 2.0 (or health in general) thus far…but as you said, I think there are now specific experimental designs that can be put into place to test empirically.
Comments (and one such experimental design proposition) here:
http://hmrx.posterous.com/calling-all-non-scientists
Hope our paths cross at some point in the future-
Jen
Joerg Kurt Wegner
August 16, 2009 at 15:34
As mentioned in my recent post did John Wilbanks (VP creative commons) made a nice statement for the difference of the IT versus life science industry “In life science we are talking about evolved systems, e.g. humans, and not about designed systems, e.g. IT chips”. In other words “open” means very different things for many people, and publishing “source code” might be very different to publishing “gene sequences” of humans. The legal aspects are much more serious, surely critical and something we should work on, but we should never forget that we must respect patient privacy for helping patients. Patients should be able to decide who is getting which information and for what. On the other hand, patients should also accept that scientists are trying to help patients by working on very complex mechanisms happening in humans, this is not easy, and hiding information is certainly not making things easier.
http://miningdrugs.blogspot.com/2009/08/web-for-patients-pets-and-scientics.html
I am strongly believing in Knowledge=People+Information, so just having people or information alone does not help, we need both, and we must work on our trust relationships ! In other words, as I will present on BioIT
http://www.bio-itworldexpoeurope.com/Bio-It_Europe_Content.aspx?id=90734
Drug Innovation 2.0 – Why we Need Knowledge Metrics for Democratic Action
Jörg Kurt Wegner, Ph.D., Scientist, Integrative Chem-/Bio-Informatics, Tibotec (J&J, Belgium); Blogger, Mining Drug Space; Project Administrator, Open Source Development
In this presentation we share three current challenges for innovation in drug design. 1) Group dynamics – Scientific collaboration typically occurs through networks. Inefficient collaboration and networking can lead to tunnel vision; innovation opportunities through more distant networks may be overlooked. Knowledge metrics can encourage new and productive networks, by appropriately rewarding contributing network members. 2) Information overload – Information and knowledge are not the same thing. We have too much (unstructured) information and we lack the time required to structure this information into knowledge. Knowledge metrics can encourage the ranking, structuring, and accessing of information on a reward-per-use basis. 3) Data silos – Data silos caused by technical or license hurdles, can reduce the number of efficient collaboration options. Knowledge metrics should clearly reward barrier-breaking and silo-bridging efforts, and favor new and diverse over redundant information…Drug designers, lawyers and computer scientists have a unique opportunity to take on these challenges – drug innovation 2.0!
viviani simpson
August 29, 2009 at 09:35
I do not agree with those people who think than Innovation is not grow.I think that Innovation is grow day by day.
http://www.goarticles.com/cgi-bin/showa.cgi?C=1839563
Marcin Wojnarski
September 23, 2009 at 15:39
Paweł,
I’ve found your blog through mloss.org. It is a nice surprise to come across someone’s blog on the net and discover that the author works 2 min. walk away from me – presumably we crossed already many times in Biology Dept where I’m a frequent visitor in the lunch time 🙂 I’m from Maths & Informatics Dept, working in the fields of data mining and machine learning – this includes occasionally bioinformatics, too.
I like your blog very much. Openness is the direction where science will go, no matter if someone likes this or not. You may take a look at recently finished Netflix Prize contest: http://www.netflixprize.com/ where $1M was awarded. Thousands of people out of academia struggled there to solve very difficult scientific / algorithmic problem and many scientific discoveries were done along the way.
I think for openness to arise two things are needed: (i) the request from the outside world for scientific innovations, with the world being ready to pay for them; and (ii) organization / infrastructure / tools / common platform – that will support openness and create a common place (like a website) where open science may live and grow. We’re trying to create something like this for data mining / machine learning – a website accompanied by a set of tools that will facilitate collaborative designing, evaluating, comparison and exchange of ML/DM algorithms. The system is called TunedIT and you may find it at:
http://tunedit.org
TunedIT is open for everyone – you can submit new algorithms, datasets and other resources no matter how many publications you’ve got in refereed journals or from what university you come. Everyone can download your algorithm and run a test to see how good it is. The result will be submitted back to TunedIT where everyone can view it and compare with results of other approaches to see which ideas are better – yours or someone else’s. So what finally matters is the objective quality of your algorithm and not your academic background.
I invite you to look at TunedIT – maybe you’ll find it useful for biology/bioinformatics research as well.
-Marcin
Jurek
October 20, 2009 at 09:37
Hi Pawel,
I am a bit surprised that you continue to be freelancing scientist in the current trend of academia to be profitable and patent research.
I am not up to date about situation in Poland – do university technicians, floorsweepers etc. are also freelancers? Did any of you professors volunteered to share his/hers own salary with you?
I suggest that you are de facto exploited by the institution you are affiliated with, and you should simply find yourself a well paying job in the industry. Possibly abroad.