HN user

lilfrost

216 karma
Posts5
Comments10
View on HN

There is consensus documenting why you did something is good (which is what root comment is talking about). Documenting what you did is commonly thought to be a crutch for writing unreadable code.

How easy it is to try this stuff out now? I remember when google first came out with the deep dream stuff I could never get it working.

I can't believe this is a published article. 3/4 of the paper is essentially crackpottery (e.g. Proposition 5 proves that the composition of surjective maps is surjective), while the final 1/4 is a "proof by picture" (see Figure 6) that you'd see from a mediocre undergraduate pset.