Open post xeroxed @xeroxed@mastodon.social · 3mo ago Replying to @liberty@mathstodon.xyz https://gracefulliberty.com/llms/ > For programming, I write code myself, or or generate it deterministically from higher-level instructions. > For writing, I use my brain and write it myself, revising it as needed. > For proofreading, I ask my friends. > For searching, I use domain-specific search tools and specific websites. > For learning, I seek out books, tutorials, and other educational material made by people. > For extra support, I ask experts in forums, chatrooms, or via email. @liberty@mathstodon.xyz For searching, I use domain-specific search tools and specific websites. want to know more above this
Open post xeroxed @xeroxed@mastodon.social · 3mo ago Replying to @tao@mathstodon.xyz <p>My talk on "New Mathematical Workflows" at Stanford last week is now online: <a href="https://www.youtube.com/watch?v=Uc2zt198U_U" target="_blank" rel="nofollow noopener" translate="no"><span class="invisible">https://www.</span><span class="">youtube.com/watch?v=Uc2zt198U_U</span><span class="invisible"></span></a></p><p>One key recommendation in the talk is to now de-prioritize the historical emphasis on competing to be the first to provide a proof for a given unsolved mathematical problem. When we were in the proof scarcity era, the "local" goal of obtaining any proof at all for a problem was fairly well aligned with the more "global" goal of collectively advancing our understanding of mathematics as a community. However, now that the ability to optimize this local goal has increased rapidly to the point of "proof abundance", we have now reached the point where Goodhardt's law <a href="https://en.wikipedia.org/wiki/Goodhart%27s_law" target="_blank" rel="nofollow noopener" translate="no"><span class="invisible">https://</span><span class="ellipsis">en.wikipedia.org/wiki/Goodhart</span><span class="invisible">%27s_law</span></a> has kicked in, and further unrestricted overoptimization of this goal will no longer create genuine mathematical progress, and may in fact inhibit it in various ways. However, there is scope for more controlled, and still meaningful, optimization in this direction along carefully chosen workflows (such as mathematics competitions) that are specifically designed to accommodate heavy AI use.</p> @tao@mathstodon.xyz sir the video is now private
Open post xeroxed @xeroxed@mastodon.social · 3mo ago Replying to @jonn_blanchard@retrochat.online <p>My wife just asked me if Github was a tech version of Pornhub and I didn't have an adequate argument that it wasn't</p> @jonn_blanchard@retrochat.online i always hesitated say the name in front of others, just feels awkward