Over the past few years we’ve seen a fascinating shift in how organizations have been managing their redundant and un-used intellectual property. For instance, it’s increasingly common for companies to open up their under-utilized patents to communities of innovators for them to do their worst with. Any products derived from the partnership see their revenue split between IP owner and innovator.
Similar moves are afoot in the scientific community, with open notebook science a movement whereby research is completely open and transparent. The aim is to ensure that unsuccessful research gets as much exposure as successful research, thus enabling academics to become smarter in the way they undergo research.
So it’s fascinating to read this week about Algorithmia. They’re a new site that have been launched to connect up organizations with lots of data with algorithms that are being under-utilized. The aim is that this match up will create not just new insights but extra profits.
The site was developed after the founders realized that an enormous amount of intellectual property was languishing within academia. Algorithms would get published in a recent paper, but then would often gather dust, despite them having the potential to solve any number of organizational problems.
“Most of them have an algorithm and software associated with them, and the problem is very few people will find them and almost nobody will use them,” the founders say.
They believe that a big part of the problem is that academics are often writing for other academics. Therefore their work is often quite challenging to decipher for the outside world, which takes up time and resources. Thus they thought the market needed a better and more efficient way of finding out this kind of information.
The process behind the site is probably a familiar one to most of you. Organizations hunt down algorithms, and if one is found that meets their requirements, the academic is paid a fee for the use of their algorithm, and Algorithmia takes a small cut. The site is currently in beta testing mode, but hopes to launch towards the end of the year for a more extensive public testing period. There is also an option to make an algorithm freely available should you wish.
Any algorithm submitted to the site has to use a standard API that makes them easier both to use and to compare against one another. Early match-ups have seen algorithms found to extract data from receipts amongst other applications.
In addition to using algorithms, users can also vote and comment on them, just as you could with any other social networking site. The site encourages users to lift the lid on their algorithms so that other users can see the code used in their construction. The hope is that the community can spot errors and suggest improvements to the code itself.
It’s a pretty neat idea that takes an innovative approach to tackling this problem. Certainly one to watch.Original post