yes, I read the first sentence, it was the point of contradiction in your other two. My friend, I don't doubt you have reasons to believe what you think, but your comment as worded is self-contradicting and lacking those reasons.
replyThey said they think extrapolation is too simplistic of a method. Then said that is why they don't agree with the pushed extrapolation of LLMs becoming SI.
replyThis is addressing the commonly repeated line of thinking that assumes LLM capability will continue to see the improvements that have been seen in the past so far.
R>C>P