Related, from the same author: <a href="https://maxwellforbes.com/posts/dont-make-things-actually-work/" rel="nofollow">https://maxwellforbes.com/posts/dont-make-things-actually-wo...</a><p>As potent and productive as the combined perspective sketched out by the author might be, it makes me wonder whether this might also point to the root of certain problems with research publishing in machine learning / artificial intelligence. I anticipate that this opinion might be polarizing, and I'm curious to hear other opinions, especially because my research background has origins in a different field.