The Bias Problem in Machine Learning

There are different kinds of requests that produce answers in which inaccuracy may matter more or less. And so that's I mean that that's kind of the bias problem. The model isn't the model is correct the model is doing what you asked it to do to your point but it's not the answer that you want. So presume for the sake of argument all of these models are 93% accurate, which is obviously nonsense but presume they're all 93%.