Artificial Intelligence tools such as Google Assistant are only as good as the information you share with it.
Artificial Intelligence at its core is automation based upon predictive analytics. Whether it is predicting the next word in a sentence, suggesting what movie or book you may like to watch or read next, predicting what wine you may enjoy with your pork-based entree, or whatever the use case, the underlying machine learning algorithms, neural networks, natural language processing algorithms and large language models all learn based upon data it is fed.
So when a Tech blogger criticizes Google’s Assistant for not recommending specific, healthy snack choices I wince. “Did you share your dietary habits with Google?” I think to myself. Does Google Assistant know that you like natural peanut butter but hate raspberries? Does it know you regularly eat egg whites, but detest tuna? How could the AI possibly create healthy snack food combinations if it does not know your dietary habits? This Tech blogger provides just one example of the many use cases I have read about where user expectations are beyond reasonable. AI platforms can’t read minds. However, if you feed them sufficient amounts of quality data, they can predict behavior and outcomes and even create new content (Gen AI), based upon provided parameters.