One part of AI is machine learning, where the machine derives expertise (= algorithms) from experience based on training data. And that's where Big Data comes-in: if the training data are indeed 'big', then our AI can learn from many many data points (aka circumstances) and that makes the AI algorithms more and more useful.
Here's one catch: since the training data are not the same, AI algorithms are not reproducible as they train themselves using the ever-growing 'big' training data.
No comments:
Post a Comment