Overindexing on the M L Space

i wanted to go back to a little bit of something you were saying before aboutke over indenxing on computer vision. Would you say for people who were more classical computer architects, it's a little bit the way we revolve around speck in the in the general compute space? So spect speck was an interesting example because we, you see, the goal was to come up with a diversity of work loats but then immediately we started to a kind of over pivot and over optomize for those speck bench marks.