The article introduces how machine learning could be used in software development. Currently, the process of software development could contain three parts of code: view, model and controller. Accordsing to the author, under the current structure, machine learning could be used in the "model" part. He points out that rather than have to decide on how everything has to work at the beginning, machine learning could help developers teach your software what to do. In that way, if developers had to change it later, they could just teach it again.
In terms of data collection, the author uses an example of pet recommendation application which tells the users which pet they should buy based on their preferences. The input would be the characteristics. The software would train the model based on the input. Unlike a database model, which has to return an exact matching dataset using complicated join statements, the application could return the top 3 most probable choices in one line of code.
Traditionally, one of the challenges in adopting A.I. was that you needed to run these models in the backend. So first of all, you needed an actual backend server, which often meant learning a different programming language, and the hassles and costs of hosting and so forth. Secondly, it meant those models could only be run when connected to the server. So if it was a core feature of your app for example, it would only work online.