Machine learning that's put into use other than creating fugly drawings no one fucking asks for about is actually incredibly useful. But it's all very boring from a consumer viewpoint that boils down to "streamlining industrial processes." It's well within reach to have an algorithm designed to manage how much paper printers within a building use up and when to add more paper so there's as little downtime as possible, but most people don't have to worry about managing 100+ printers. I saw a short clip of some Chinese port using ML to streamline port operations. The algorithms are set to recognize what a shipping container is and where to load and unload the container. Since the containers are standardized, it's not hard to train the algorithms with past data.
Machine learning that's put into use other than creating fugly drawings no one fucking asks for about is actually incredibly useful. But it's all very boring from a consumer viewpoint that boils down to "streamlining industrial processes." It's well within reach to have an algorithm designed to manage how much paper printers within a building use up and when to add more paper so there's as little downtime as possible, but most people don't have to worry about managing 100+ printers. I saw a short clip of some Chinese port using ML to streamline port operations. The algorithms are set to recognize what a shipping container is and where to load and unload the container. Since the containers are standardized, it's not hard to train the algorithms with past data.