In pretty much every single HN post ~~on this topic~~, there are a number of commenters that are probably bots. Anyways it makes sense from a theoretical standpoint that LLMs should be able to find flaws in code better/faster than humans eventually, and its reasonable to think that time has come