1. 3
  1.  

  2. 4

    I read this as “can you fit a whole genome into a QR code”. Weird.

    For anyone wanting an answer to that question, here’s a SO question and answers:

    https://bioinformatics.stackexchange.com/questions/1/whats-the-most-efficient-file-format-for-the-storage-of-dna-sequences/20

    1. 1

      It is nice, but the statement that human has always capabilities to write faster code in assembly than compiler from C is utter bullshit. In theory - maybe, but in reality compilers optimisations can reason about the code way beyond human capabilities.

      1. 4

        Not beyond human capabilities, just usually beyond human patience.

        1. 1

          Let’s assume you are right.

          Let Z be the program optimized better by a compiler than a human.

          For Z to exist, there must be a logical or computational structure in the program that no human world recognize, but the compiler would. We assume such a structure exists.

          However the compiler only does optimizations based on rules. Thus for Z to exist, there must be a rule a human can not understand.

          Is there any rule written in to a compiler that a human can not understand? I will posit thus: a compiler rule is written or verified by a human. Thus for any given rule, there exists a human who can understand it and is able to calculate its correctness themselves.

          Thus there is no rule beyond human comprehension.

          There is no computational or logical structure in a program that a human can not parse as well as the optimizing compiler.

          And a human optimised program can always be equally fast as a compiler optimized program. Z does not exist, and you are not correct.