Post B9V58WpLi7Wbz2wbUu by Pepijn@mastodon.online
 (DIR) More posts by Pepijn@mastodon.online
 (DIR) Post #B9V58WpLi7Wbz2wbUu by Pepijn@mastodon.online
       0 likes, 1 repeats
       
       I'm reading a textbook authored by an A. Lesk. Illustrating the dog genome section is a lovely drawing by a V.E. Lesk. It's safe to assume the two Lesks are related.Most of the pages are text only. And while there's little anecdotes it's primarily a classic "information-first" work.Yet... the dog and the overall finesse in the text gives an undeniable feeling of a human created / owned work.I wonder how far we're away from LLM to replicate that for this type of texts. Is it even possible?
       
 (DIR) Post #B9V58XXf3MyECTfzZA by Pepijn@mastodon.online
       0 likes, 0 repeats
       
       It's amazing how easily I find myself distracted by information dense reads. Wish I remembered how I handled this at age 20.Take this page where extinct dork-birds from New Zealand are used to illustrate palaeosequencing. So either Emeus and Euryapteryx looked exactly the same, or someone (by accident?) reused an image.The error (if so) is at least not in the textbook as the original paper has the same image. The textbook does have an error in the doi / pnas link (https://www.pnas.org/doi/epdf/10.1073/pnas.0906660106)