Grok was ask a pretty simple question and was surprised that it spit out fake news.

Grok could have looked at all available public data about how many Aircraft were shot down in the first 20 days of Operation Iraqi Freedom, but instead it decided to make up facts and said zero fixed-wing Aircraft were lost when this isn't what happened.
Now that AI is everywhere how are people supposed to trust the outputs? It also brings up an interesting thing about trusting data you get from websites and this is something @CommonCrawl is doing right now.
They have the Largest archive of the internet and used by 80% of LLMs. They pull in recent web crawls and hash it on-chain as an Immutable record. AI developers now have a higher confidence of their data. This archivable data can prove the data wasn't tampered.