I have been researching MEV these days and unexpectedly discovered that several LLMs still have many bugs when doing numerical statistics Many of the principles of MEV are basically correct when asked, but when asked about the data, Claude's set of 2024 ETH MEV data caught my attention and intuitively appeared to be on the high side I followed the link it provided and found that there was indeed a problem - the original article stated - MEV Revenue Ethereum added $50M (an increase of 50M throughout the year). Claude actually told me the monthly MEV revenue of 50M Driven by curiosity, I threw this article to GPT to do a 2024 MEV revenue analysis. However, GPT made a similar mistake as Claude - it did not mistake the 50M growth, but instead treated the 60M revenue for the following quarter as monthly revenue. Later, I questioned it, and it admitted to it- The mistake I made was: Put '$60M (quarterly)' Misunderstanding has become $60M per month This is typical: Misalignment of quarterly as monthly ” Then we continued to test Gemini and Grok, and Gemini answered correctly, while Grok was the most speechless- According to the article link you provided（ https://sparkco.ai/blog/ethereum ）The article does not explicitly mention the specific total MEV revenue of ETH in 2024. Completely ignored the ETH quarterly revenue chart (including MEV) in the article When you use lobster employees to do this statistics, you'd better let them grow Dim sum. It's better to cross check two or three brains