Probably most readers won't actually use this directly, but their tools will. In a nutshell, this is a directory of AI models that can be used for information processing by other applications. For example, one of the models listed is the Toxic Comment Classifier. This is a pre-trained neural network that detects six types of toxicity in a text fragment. To use it, you would set up an instance (by executing a one-line Docker command), and then send the text fragment to the instance API from your own application. This gives your application a way to detect toxic comments, and then you can decide whether to delete them or whatever. The point here is that this is accessible as a service, and that applications of the future will access a large number of these services to perform cognitive tasks on demand.
Today: 4 Total: 94 [Direct link] [Share]
