Comment by red75prime
3 hours ago
Their diagonalization argument applies to any system that uses finite training data. Calling such a system "LLM" is an (unintentional) red herring.
3 hours ago
Their diagonalization argument applies to any system that uses finite training data. Calling such a system "LLM" is an (unintentional) red herring.
Yeah. IMHO this is the more serious objection.