Reddit user reports Claude replacing precise terms with random synonyms
A Reddit user reports that Claude models sometimes replace established terminology with random synonyms, particularly in mathematical and code-related work. The user cites an example in which "morph the routine" became "reshape the routine," arguing that such substitutions can alter code-related meaning and confuse a subsequent agent. The user also hypothesizes that watermark-related changes may have increased output entropy, but provides no confirmation of that causal theory.
The report highlights a potentially consequential weakness in Claude output: semantic substitutions may reduce readability and introduce ambiguity in code or plans passed to another agent.
It is an extreme productivity drain to have these models substitute settled vocabulary with random synonyms, particularly when working on math or code.
No, Claude, you cannot just substitute "morph the routine" with "reshape the routine" because only Morph is in the code while reshape could literally be anything.
This semantic substitution makes it extremely tedious to understand its work and output and, moreover, occasionally cause it to write a wrong plan that gets interpreted in a different way by the next agent picking up the plan.
I almost suspect that this is a result of needing to satisfy watermarking regulation. Now, I'm not saying that watermarking itself caused this. I am instead hypothesizing that as part of a project to satisfy regulation, Anthropic created more semantic substitution in their models.
Now, as for why I have this hypothesis: