Does AI make Lazybones?

We have heard a lot already about Artificial Intelligence (AI), and in the coming years this is surely only going to grow. I am not talking about the AI as in Kubrick’s 2001: A Space Odyssey where HAL (the name looks a lot like another three letter computer giant) refuses to open the pod bay doors I am also not talking about the AI in self driving cars that have to decide on the trolley problem. No, I am talking about everyday AI that makes life easier by taking simple and mundane tasks and helping us with it. One of the tasks that AI, and in this case Deep Learning to be exact, can surely help with it the classification of photo and video material. Services like ClarifAI can automatically help to automatically tag what is in a certain photo without human intervention. Just show the system enough examples with what is visible and you build up something called the Ground Truth. An example: show enough pictures of weddings and the system can figure out the setup of the wedding: determining who is the bride, groom et cetera. Helpful or only a nice to have for lazybones?

Instapeople

Instant coffee is seen by many people as being not really coffee. I agree with that sentiment. There are certainly situations where instant coffee is handy, but I prefer my coffee being brewed on the spot. Instagram, the popular photo and video-sharing social networking service, is a fun way to share things with your friends. One of the objections to the use of Instagram by its users, is that it portraits a world that is too perfect to the extent of almost being fake. If you do not know what I mean, read this post about what Popular Instagram blogger Chessie King does to combat that.

When a colleague recently told me about a new AI tool to recommend engaging captions and perfect hashtags to assist influencers and brands with social media posts, I almost peed my pants. The AI model (using deep learning) is trained to create captions in the parlance of an Instagram user. Where as normal AI model would recognize a photo of for instance a street in Manhattan with cars, people and buildings, the new AI tool creates the result: “Love this view! Life in one of my favorite cities.”. When this is created, it is the ultimate level of fake. How do you infer that information from the photo? Do not get me wrong, I am a big fan of AI with of course a portion of caution. But this is really silly stuff. It is ok to make things easier but when AI is actually creating captions like that, you are simply too lazy.

APIs and Microservices

All of these API capabilities are often opened up by using API technology like the one we use WSO2 Micro Gateway and WSO2 Micro ESB to name a few. APIs are also a way of making things easier. With a simple URL, information can be retrieved way easier than using soap services. We are hiding complexity more and more and making integration easier, especially with services like our own Connext ‘Managed Middleware as a Service’ platform which is the ultimate WSO2 Cloud Service. Not so much using a one-size-fits-all approach but a fit-for-purpose approach. In some situations, you have an Enterprise Integration solution in place. Other cases warrant an API Management solution and in the future, for services that have a high demand, we will see a Microservice approach using a containerized environment to scale with expected demand.  If you want to know more about Microservices, read our WSO2 Microservices white paper!