```todoist name: Today filter: "Today | Overdue" ``` Clearly, such a “feature” could never be de- tected on a given occasion; yet “adjusting” to it in absentia, as a means to correct reidentifications, would be of great value. Accordingly, *whatever incorporated the ability to recognize those faces could, by our account, be deemed a representation* of that feature. To be sure, that initial account was formulated only casually and without argument, so nothing is proved by it; still, the consilience is not unsuggestive. 1. I would like you to try that again. This time, never say anything is important, this is not how a real academic would write. I wpould also like you to make your changes/additions slightly later in the text: after "in a similar manner, records textual data from the internet through activation weights and represent the ability to recognize input texts and react with the proper output texts through connection weights." and before "In sum, the same neural [[network that]]," Also, make sure that first you raise [[the question]] of whether it is real know how or just seems like it is. Then talk about what [[Haugeland]] says about the ping pong robot (using the indicative quote to show his view). Then explain why what he says about the ping pong robots havining know how can be applied to the case of LLMs' discursive know how. 2\. Next, something completely different. I don't want you to make any more changes to my text just yet, but i want to get your opinion on whether discurssive knowledge how, as we have defined it in the paper call be thought of as bullshitting according to frankfurt. I have attaced Frankfurt's book on the subject. Please quote from the book when giving me your answer. https://youtu.be/PTruNkpKmZc?si=VLQwScsk5MKm8fVW