What I take Leibniz to have meant was that when he uses math he is much less prone to self-deception and to mistakenly believing he’s had an insight than when he uses natural language, so he tried (and failed) to extend math so that he could use it to talk about or think about all of the things he uses language to talk about, including human and personal things.
Gottlob Frege, the creator of predicate logic, had a similar ambition.
Note that creating FAI that will extrapolate the volition of the humans requires using math (broadly construed) or formal language to talk about some human things. In particular, you must formally define “human”, “volition” and the extrapolation process. The fact that Leibniz and Frege did not get very far with their ambition (although the creation of predicate logic strikes me as some progress) suggests that for us to teach ourselves how to do that might require nontrivial effort—although I tend to think that we have a head start in some of our mathematical tools. In particular the AIXI formalism (and to a lesser extent) some of the more intellectually-deep traditions we have for designing programming languages and writing programs strike me as superior to any of the “head starts” (including predicate logic) that Leibniz or Frege (who died in 1925) had at their disposal.
(Pearl’s technical explanation of causality is another things that sort of seems to me that it might possibly somehow assist in this enterprise.)
SIAI has not included me in their private or not-completely-public discussions of Friendliness theory to any significant degree, so they might have insights that render my speculations here obsolete.
Another person who seems to have had the same general ambition as Liebniz and Frege is the Free Software Foundation’s lawyer, the man who with Richard Stallman created the General Public License. Eben Moglen. Here’s Moglen in 2000:
I was committed to the idea that what we were doing with computers was making languages that were better than natural languages for procedural thought. The idea was to do for whole ranges of human thinking what mathematics has been doing for thousands of years in the quantitative arrangement of knowledge, and to help people think in more precise and clear ways.
What I take Leibniz to have meant was that when he uses math he is much less prone to self-deception and to mistakenly believing he’s had an insight than when he uses natural language, so he tried (and failed) to extend math so that he could use it to talk about or think about all of the things he uses language to talk about, including human and personal things.
Gottlob Frege, the creator of predicate logic, had a similar ambition.
Note that creating FAI that will extrapolate the volition of the humans requires using math (broadly construed) or formal language to talk about some human things. In particular, you must formally define “human”, “volition” and the extrapolation process. The fact that Leibniz and Frege did not get very far with their ambition (although the creation of predicate logic strikes me as some progress) suggests that for us to teach ourselves how to do that might require nontrivial effort—although I tend to think that we have a head start in some of our mathematical tools. In particular the AIXI formalism (and to a lesser extent) some of the more intellectually-deep traditions we have for designing programming languages and writing programs strike me as superior to any of the “head starts” (including predicate logic) that Leibniz or Frege (who died in 1925) had at their disposal.
(Pearl’s technical explanation of causality is another things that sort of seems to me that it might possibly somehow assist in this enterprise.)
SIAI has not included me in their private or not-completely-public discussions of Friendliness theory to any significant degree, so they might have insights that render my speculations here obsolete.
Another person who seems to have had the same general ambition as Liebniz and Frege is the Free Software Foundation’s lawyer, the man who with Richard Stallman created the General Public License. Eben Moglen. Here’s Moglen in 2000: