I wanted to give a rebuttal with all the logical fallacies in the Memo, but I tried to let LLMs do the work first, as the memo suggests. So, while disagreeing with the memo, I followed its advice successfully. Yet it seems that Tobi himself didn't listen to his own advice and didn't try to let an LLM do the memo instead.
https://chatgpt.com/share/67f56ffe-b408-800e-ad76-46a12eabe0...
Besides all this, Point 5 is truly bonkers. How I read it is an unintended epistemological farce. The null hypothesis is that AI can do your job. Then they double-down on their burden-of-proof horsing by asking you to prove your work can't be done by a hypothetical, idealized AI.
I wanted to give a rebuttal with all the logical fallacies in the Memo, but I tried to let LLMs do the work first, as the memo suggests. So, while disagreeing with the memo, I followed its advice successfully. Yet it seems that Tobi himself didn't listen to his own advice and didn't try to let an LLM do the memo instead. https://chatgpt.com/share/67f56ffe-b408-800e-ad76-46a12eabe0...
Besides all this, Point 5 is truly bonkers. How I read it is an unintended epistemological farce. The null hypothesis is that AI can do your job. Then they double-down on their burden-of-proof horsing by asking you to prove your work can't be done by a hypothetical, idealized AI.