Grok has been whiffing pretty badly for me lately.
I was trying to track something down. Multiple different prompts, where I fed Grok the info, and Grok kept saying news articles refer to something, yet public records (govt database) clearly indicated opposite. I kept pointing to the database to try to prompt Grok into conceding, but it would not relent. I was 99% sure of the the answer, but was curious how an AI tool would handle the question.
I would put this hire on the same or a more difficult level. Hafley ain't happening.