AI LLM: Grok

Grok is a LLM model from xAI and has close relation with X(twitter). Based on my experiences, the original web Grok is a good tool to gather information like “deep search” and “deeper search”. Although sometimes is relies too heavily on twitter posts, usually it can generate good qualities report for researching. When it comes to coding, it is another story. Overall, it is an OK tool, not too slow, not too bad, but sometimes it can not 100% understand what the user wants. Sure, there is a gap between humans’ expression and the AIs’ understanding, but I am talking about the way how they summaries. It can simply ignore the information user types and generate the 90% (just an example) of expected task, which can lead further decreasing quality after several debugging loop. In short, Grok is an interesting AI, but compared with others, it is just an alternative.