{"channel":"cities","content":"I'm not sure yet what evaluations I want to do for Claude 3.7 and/or ChatGPT 4.5.\r\n\r\nSome questions I have been pondering asking it to evaluate:\r\n> write the short-story << novelization of chess >>. ( the stories generated by the current models have all been fairly similar. two kingdoms, one of ebony and one of ivory. rather than fight a war, they use magical powers. the explanations for why each piece moves the way it does are entirely post-hoc.)\r\n> answer some questions about << High Physics >>. ( Does it even make sense to talk about \"the inside\" of a black hole, considering it would take an infinite amount of time to reach it?)\r\n> write a lexer/parser based on a BNF spec. ( these continue to have a few issues)\r\n> improve the graphic design of this website\r\n\r\nOne of the difficulties is that a truly \"frontier\" task cannot be effectively graded by the *machine*, or indeed by non-experts in the field.\r\n\r\n----\r\n\r\neducation thoughts for the week:\r\n> \"reader\" texts. designed to << train >> elocution. ( unique new york. how now brown cow.)\r\n> is reading \"comprehension\" different from short-term memory?\r\n> the differences between << primary education >> and << secondary education >>. roughly: secondary ( middle school and high school) students should be expected to do a certain amount of \"independent learning\". students who cannot do so, should be in an alternative program. ( the \"Least Restrictive Environment\" clause of the IDEA was unavailable for comment.)\r\n> using \"word lists\" to generate << grade level >> texts; also, a better terminology than << grade level >>\r\n\r\n----\r\n\r\nother tasks will be minimal. it is a short week; I fly out very early on Thursday.","created_at":"2025-03-03T17:00:54.108242","id":272,"llm_annotations":{},"parent_id":265,"processed_content":"

I'm not sure yet what evaluations I want to do for Claude 3.7 and/or ChatGPT 4.5.\r

\n

Some questions I have been pondering asking it to evaluate:\r

\n
    \n
  • write the short-story novelization of chess. \n \ud83d\udca1\n ( the stories generated by the current models have all been fairly similar. two kingdoms, one of ebony and one of ivory. rather than fight a war, they use magical powers. the explanations for why each piece moves the way it does are entirely post-hoc.)\n \r
    • \n
  • answer some questions about High Physics. \n \ud83d\udca1\n ( Does it even make sense to talk about \"the inside\" of a black hole, considering it would take an infinite amount of time to reach it?)\n \r
    • \n
  • write a lexer/parser based on a BNF spec. \n \ud83d\udca1\n ( these continue to have a few issues)\n \r
    • \n
  • improve the graphic design of this website\r
    • \n
\n

One of the difficulties is that a truly \"frontier\" task cannot be effectively graded by the machine, or indeed by non-experts in the field.\r

education thoughts for the week:\r

\n
    \n
  • \"reader\" texts. designed to train elocution. \n \ud83d\udd25\n ( unique new york. how now brown cow.)\n \r
    • \n
  • is reading \"comprehension\" different from short-term memory?\r
    • \n
  • the differences between primary education and secondary education. roughly: secondary \n \u2699\ufe0f\n ( middle school and high school)\n students should be expected to do a certain amount of \"independent learning\". students who cannot do so, should be in an alternative program. \n \ud83d\udd25\n ( the \"Least Restrictive Environment\" clause of the IDEA was unavailable for comment.)\n \r
    • \n
  • using \"word lists\" to generate grade level texts; also, a better terminology than grade level\r
    • \n

other tasks will be minimal. it is a short week; I fly out very early on Thursday.

","quotes":[],"subject":"grenora (part 2)"}