I don't think it's a big leap to assume that people running bots review the output less and less before acting on it.
I don't think it's a big leap to assume that people running bots review the output less and less before acting on it.