I don't know that we're in the singularity, but we are certainly past the point where LLMs can only write spaghetti code. LLMs can write complex systems, when properly lead by software engineers. They can produce code faster, and at a higher quality, than purely human endeavors.

The higher quality part is the part people are missing. You can write much more robust code using LLMs because you can employ more comprehensive testing strategies. People are using LLMs to find hundreds of vulnerabilities in popular software. Imagine how much more secure software can be when LLMs because integrated into the process of writing, testing, and penetration testing code.

I think it might historically be defined as the point where critical thinking was replaced with blind deferAInce