Lukas
Jacob Something@darkcisum Apart from the fact that you'd use nonlinear stuff inside that mix too, it is precisely to the point.
@@darkcisum Continued image description:
The dialog is
A: _This_ is your machine learning System?
B: Yup! You pour the data into this big pile of linear algebra, then collect the answers on the other side.
A: What if the answers are wrong?
B: Just stir the pile until they start looking right.
Leonie 
@darkcisum stupid people might also call that ai
llewelly@darkcisum almost 7 years old, and it explained it perffectly.
Elias> almost 7 years old
I think a difference now compared to 7 years ago is the size of the pile, it's gone from millions, to billions, to trillions of parameters that are all affected by the "training" data in ways that are more inexplicable than ever.
FiXato@darkcisum "See, people of the time misunderstood the term "machine learning"; thinking it meant the machine was learning, while in reality the true exercise was to learn that the machine was often wrong."
— some future historian
— some future historian
myrmepropagandist