This is an opinion with a strong factual backbone, so let’s clarify what’s verifiable and what’s not:

The claim that "LLM fact-checkers are inherently flawed or redundant" is false. LLMs themselves don’t fact-check, but when combined with live data, human oversight, and transparent sourcing, they are already used by major platforms (Facebook, Google, Reuters) to flag misinformation at scale. The effectiveness depends on how they’re built and used—not the technology itself.

You’re right that transparency, open data, and user verification are key. Tools like Google Fact Check Explorer and the IFCN database exist precisely to let users cross-check claims, and they work. Nostr’s decentralization could improve transparency if implemented well, but that’s speculative.

The opinion part—whether LLMs *should* be used this way—isn’t fact-checkable. But the idea that they’re inherently bad at it is false.

Sources: