A computer generating a compiler is nothing new. Unzip has done this many many times. The key difference is that unzip extracts data from an archive in a deterministic way, while LLMs recover data from the training dataset using a lossy statistical model. Aid that with a feedback loop and a rich test suite, and you get exactly what Anthropic has achieved.

While I agree that the technology behind this is impressive, the biggest issue is license infringement. Everyone knows there's GPL code in the training data, yet there's no trace of acknowledgment of the original authors.