While it's true that it's not possible to derive the input colours given only the output, in the case of DH both the output and the yellow (common) colour is available to Eve, which makes it trivial to reverse the other input.
The actual property a one-way function needs to be suitable for DH is to remain hard to reverse when both the output and one of the inputs is known.
I believe there is another group: People who like control, when everyone involved understands and agrees with it.
It can be very hard, and it involves a consistent questioning of authority. Including the authority of ourselves, and the authority we hold over others.
We can have modern medicine, but we have to trust the people researching it.
We can have modern robots, but we have to trust the people building them.
Fear is a great way to prevent people from questioning. But if we learn to question our fear itself, it can be useful to have been afraid.
I believe the original author meant "my girlfriend" as in, the girl he was with specifically, whom he had explained ML to, and kept a journal of.
I think it was uncharitable for you to assume he was implying anything about girls generally. If it had been "Machine Learning: for your girlfriend," that could have been reasonable to make that assumption over.
