# What Machine Learning can and cannot do - [[Supervised Learning (SL)]] - Can do pretty much anything we can do with a second of thought can be done/automated with Supervised Learning - Can perform complex tasks that take hours/days or longer to perform for a human (e.g., market research) * Difficult for current AIs to map pictures/videos of humans/animals as input and identifying their intentions. It's also a difficult problem for humans * ML tends to work poorly for learning "complex" concepts from small amounts of data and for new types of data. Humans are much better at generalizing and applying previously gained knowledge to learn how to deal with new data/problem types * [[What makes ML problems easier]]