Why is Grok not in their benchmarks? I don't see comparisons to Grok in any recent announcements about models. In fact, I see practically no discussion of Grok on HN or anywhere except Twitter in general.