Machine Learning models on GCP have helped many companies to make a breakthrough in their industries.  The secret to takeaway is that ML can automate activities that can save a human team or support them. The new models have also outperformed humans in some domains recently. More than just binary classification tools, the picture classification as a specialization is also. You can play with the pre-built Vision API model which allows you to move through a JSON request and get back a ranked list of associated image labels. Even if you have more than one subject in a photograph, you can draw bounding boxes and even identify pieces inside an image.

Modern models of image classification can also create captions that illustrate what is happening in the image, like a map of dependencies, or two hockey players fighting over a puck. It is important to point out here that even the best models in their predictions could and will make mistakes. Like the road sign captioned as a fridge filled with plenty of food and drink. As you have noticed, these days there have been a lot of interesting applications for machine learning. Including tracking objects and pictures, helping with disease detection, and even allowing cars to drive themselves. But AI too can be used in more playful ways.