you are viewing a single comment's thread
view the rest of the comments
[–] 7 points 22 hours ago* (last edited 21 hours ago) (3 children)

Because the proof as presented is a Lean program. Lean is a proof checking program, that means that this type of code runs only if the logic used is correct.

and how do we know that the lean (i am hearing the word for the first time - this is not a sarcasm, but genuine question) is correct, in its design, lack of logical flaws and the implementation?

lot of the software where the exploit was found was considered reliable, because it worked on known and proven mathematical principle, until the exploit was found.

  • source
  • parent
  • hideshow 3 child comments
  • [–] 6 points 20 hours ago

    The exported raw proof has been checked by several proof checkers, including Lean’s official kernel, the con-leche and con-ron checkers, and several independent checkers contributed by third parties and implemented in different programming languages. The checkers con-leche and con-ron come with a formal proof, relative to set theory, that they cannot accept a false proof, thus addressing concerns that Lean’s logic itself might have flaws that the proof could exploit. An extensive overview is available at https://math.mit.edu/~drew/grh/. These checks ran on an inert export of the proof, out of reach of hypothetical malicious code in the Lean proof sources.

    Valid point, but we should be able to trust something at some point I guess.

    disclaimer; I have no acting knowledge of anything here.

  • source
  • parent
  • [–] 4 points 16 hours ago

    Here my knowledge gets hazy, so I am just reporting what I understood. The “validator” of Lean got build from the ground up, each step validated with a proof checker. I am unsure of the details. I know that there are different ways to compile a Lean code, some are considered “most likely correct” and some are “most definitely absolutely correct” and there have been cases that, by mistake, exploited the difference.

  • source
  • parent
  • load more comments (-1 replies)