No, they cannot. LLMs are LANGUAGE models, not KNOWLDEGE models. They do not "know" what algorithms mean any more than they know what everything else they "talk about" mean.
What they are more than capable of is to produce really convincing hype, which claims that LLMs can do anything and everything.
