On a side note Generalization isn’t bad. Entire field of AI relies on generalisation
And AI has an average IQ
For now. I’m saying taking a small training set and extrapolating isn’t a bad idea. Statistics uses this trick too.
And it’s almost never accurate to represent the entire population , which statisticians put effort in clarifying
Statement about “all” is wrong. Say if you measure average height from a small sample. It will be close to the total populations average height with high probability.
Not quite 🤣🤣
It is quite accurate. The AI relies on minimizing Genaralization error. It’s not bad as it stands now.
