There is room for something like RAISE - Revision history

Issa at 18:34, 18 July 2021

2021-07-18T18:34:06Z

Issa at 02:38, 28 March 2021

2021-03-28T02:38:50Z

Issa at 01:37, 1 December 2020

2020-12-01T01:37:57Z

Issa at 22:28, 9 November 2020

2020-11-09T22:28:46Z

Issa at 00:01, 7 June 2020

2020-06-07T00:01:21Z

Issa at 22:35, 29 May 2020

2020-05-29T22:35:29Z

Issa at 22:34, 29 May 2020

2020-05-29T22:34:48Z

Issa at 22:31, 29 May 2020

2020-05-29T22:31:35Z

Issa at 00:24, 19 May 2020

2020-05-19T00:24:15Z

Issa at 00:22, 19 May 2020

2020-05-19T00:22:25Z

@@ Line 19: / Line 19: @@
 <references/>
 [[Category:AI safety meta]]
 [[Category:RAISE]]
 [[Category:Claim]]

← Older revision		Revision as of 02:38, 28 March 2021
Line 22:		Line 22:
	[[Category:AI safety meta]]		[[Category:AI safety meta]]
	[[Category:RAISE]]		[[Category:RAISE]]
		+	[[Category:Claim]]

@@ Line 6: / Line 6: @@
 * Detailed solutions for all of the prerequisite math books, e.g. for the ones listed at [https://intelligence.org/research-guide/]. I've started on one example of this at [https://taoanalysis.wordpress.com/] (though I'm writing that blog for other reasons as well). You might wonder, why not Stack Exchange or Quora or something? I already do this, but [[online question-answering services are unreliable]], and [[Unreliability of online question-answering services makes it emotionally taxing to write up questions|this unreliability makes it emotionally taxing to write up questions]].
-* A network of tutors or people who have already worked through a particular book, where you can ask them questions in a really ''low friction'' and ''high probability of getting a response'' way. A minimal implementation of this is to have a single tutor focusing on training/helping AI safety people, e.g. tutoring to get them quickly up to speed in some undergraduate subfield, or helping them digest "[[Logical Induction]]". this requires a kind of ADHD/"living library" mindset.
+* A network of tutors or people who have already worked through a particular book, where you can ask them questions in a really ''low friction'' and ''high probability of getting a response'' way. A minimal implementation of this is to have a single tutor focusing on training/helping AI safety people, e.g. tutoring to get them quickly up to speed in some undergraduate subfield, or helping them digest "[[Logical Induction]]". this requires a kind of ADHD/"living library" mindset. See also [[Asynchronous support]]
 * Grading network for AI safety people self-studying math. It would be great if people could "own" specific problems of specific math textbooks, and be a contact person for that problem. Then, if you get stuck on something, you can email or somehow contact them to get (1) a model solution, (2) grading/checking over your proof, (3) discussions about the problem, about interesting near-by ideas, etc.
 * Writing up actually good explanations for things like [[Solomonoff induction]], [[belief propagation]], [[Markov chain Monte Carlo]], etc. Belief propagation in Pearl's book is actually ok (except for the [https://machinelearning.subwiki.org/wiki/User:IssaRice/Type_checking_Pearl%27s_belief_propagation_notation horrible notation]), but as [[Abram]] [https://www.lesswrong.com/posts/tp4rEtQqRshPavZsr/learn-bayes-nets says] it doesn't really tell you all the connections you need to know for rationality/AI safety stuff.

← Older revision		Revision as of 22:28, 9 November 2020
Line 1:		Line 1:
	Self-studying all of the technical prerequisites for [[technical AI safety research]] is hard. The most that people new to the field get is a list of textbooks. I think there is room for something like what [[RAISE]] was trying to become: some sort of community/detailed resource/support structure/etc for people studying this stuff.		Self-studying all of the technical prerequisites for [[technical AI safety research]] is hard. The most that people new to the field get is a list of textbooks. I think there is room for something like what [[RAISE]] was trying to become: some sort of community/detailed resource/support structure/etc for people studying this stuff.
		+
		+	Reasons for pessimism: If hiring capacity is limited at AI safety orgs and mainstream AI orgs only want to hire ML PhDs then new people entering the field will basically get all the necessary training in school.

	Here are some more concrete ideas:		Here are some more concrete ideas:

@@ Line 5: / Line 5: @@
 * Detailed solutions for all of the prerequisite math books, e.g. for the ones listed at [https://intelligence.org/research-guide/]. I've started on one example of this at [https://taoanalysis.wordpress.com/] (though I'm writing that blog for other reasons as well). You might wonder, why not Stack Exchange or Quora or something? I already do this, but [[online question-answering services are unreliable]], and [[Unreliability of online question-answering services makes it emotionally taxing to write up questions|this unreliability makes it emotionally taxing to write up questions]].
 * A network of tutors or people who have already worked through a particular book, where you can ask them questions in a really ''low friction'' and ''high probability of getting a response'' way. A minimal implementation of this is to have a single tutor focusing on training/helping AI safety people, e.g. tutoring to get them quickly up to speed in some undergraduate subfield, or helping them digest "[[Logical Induction]]". this requires a kind of ADHD/"living library" mindset.
 * Writing up actually good explanations for things like [[Solomonoff induction]], [[belief propagation]], [[Markov chain Monte Carlo]], etc. Belief propagation in Pearl's book is actually ok (except for the [https://machinelearning.subwiki.org/wiki/User:IssaRice/Type_checking_Pearl%27s_belief_propagation_notation horrible notation]), but as [[Abram]] [https://www.lesswrong.com/posts/tp4rEtQqRshPavZsr/learn-bayes-nets says] it doesn't really tell you all the connections you need to know for rationality/AI safety stuff.
 * Redpilling people about [[spaced repetition]] and other effective learning techniques.

@@ Line 3: / Line 3: @@
 Here are some more concrete ideas:
-* Detailed solutions for all of the prerequisite math books, e.g. for the ones listed at [https://intelligence.org/research-guide/]. I've started on one example of this at [https://taoanalysis.wordpress.com/] (though I'm writing that blog for other reasons as well). You might wonder, why not Stack Exchange or Quora or something? I already do this, but [[online question-answering services are unreliable]]. and actually this feeling of uncertainty, that I could spend time writing up my question only to be completely ignored, is one of the big reasons why I don't post more questions.
+* Detailed solutions for all of the prerequisite math books, e.g. for the ones listed at [https://intelligence.org/research-guide/]. I've started on one example of this at [https://taoanalysis.wordpress.com/] (though I'm writing that blog for other reasons as well). You might wonder, why not Stack Exchange or Quora or something? I already do this, but [[online question-answering services are unreliable]], and this [[unreliability of online question-answering services makes it emotionally taxing to write up questions]].
 * A network of tutors or people who have already worked through a particular book, where you can ask them questions in a really ''low friction'' and ''high probability of getting a response'' way. A minimal implementation of this is to have a single tutor focusing on training/helping AI safety people, e.g. tutoring to get them quickly up to speed in some undergraduate subfield, or helping them digest "[[Logical Induction]]". this requires a kind of ADHD/"living library" mindset.
 * Writing up actually good explanations for things like [[Solomonoff induction]], [[belief propagation]], [[Markov chain Monte Carlo]], etc. Belief propagation in Pearl's book is actually ok (except for the [https://machinelearning.subwiki.org/wiki/User:IssaRice/Type_checking_Pearl%27s_belief_propagation_notation horrible notation]), but as [[Abram]] [https://www.lesswrong.com/posts/tp4rEtQqRshPavZsr/learn-bayes-nets says] it doesn't really tell you all the connections you need to know for rationality/AI safety stuff.