As we build technology into everything, creating entertainment, recommendation services and other applications that can deliver whatever we need at the moment we ask for it, a new breed of application is being born. Maybe it’s the rise of the federated web. Maybe it’s a network or data; but whatever we call it, the currency of this new breed of application is the application programming interface.
APIs between services are the key, packaging up data generated by one web app and transferring it to another. In chats with entrepreneurs at South By Southwest, the API has become a basic building block, but what seems relatively unexplored is who has access to that building block and the capricious nature of the rules surrounding it. As Twitter showed those using its API last week when it gave notice to app developers that it doesn’t want new Twitter clients, sometimes applications might be building their services on a foundation of sand.
This weekend at SXSW, I had an entrepreneur named Damian Kimmelman, the founder of DueDil explain to me that he was using an API to gather some data from his friend’s company “because he liked their algorithms.” Unsaid was the fact that the guy providing the API was also his friend, but surely that ranked as well. Meanwhile in another conversation, Alex Housley, COO of Rummble Labs, which provides recommendations for businesses, is working with GNIP in part because it has access to Twitter’s firehouse of tweets from an API.
But as the Twitter example shows, there are issues associated with APIs that may mirror issues associated intellectual property licensing. Sam Ramji, of Apigee, which acts as a monitoring and quality assurance service for APIs, agrees. In a conversation at SXSW, he said the technology is there to share content, APIs, and whatever else someone wants, but the contracts and legal system will have to catch up. For example, do the clients using the Twitter API have any legal resource after the microblogging company changed the terms on them? Should they? In some areas, such as the ownership of movies for example, the contracts and business models are in place. But for APIs, the general consensus is that open APIs are good, and enticing developers to use them is only going to be good for those companies seeking to build a platform.
However, as those platforms seek to expand and monetize their user base and their existing data, and as we build out the federated software I described after hanging out at SXSW last year, the issue of API contracts or dependability will have to come to the fore in order to protect those building a business on top of such platforms as well as the users who come to depend on them. Are there people working on this, or are we still at the caveat emptor stage? I welcome any thoughts on this issue going forward.