Iterated amplification - Revision history

Issa at 03:58, 26 April 2020

2020-04-26T03:58:12Z

Issa at 01:08, 25 February 2020

2020-02-25T01:08:50Z

Issa at 22:47, 24 February 2020

2020-02-24T22:47:47Z

Issa at 22:47, 24 February 2020

2020-02-24T22:47:02Z

Issa at 22:41, 22 February 2020

2020-02-22T22:41:21Z

Issa: Created page with "'''Iterated amplification''' (also called '''iterated distillation and amplification''', and abbreviated '''IDA''') is the technical alignment agenda that Paul Christiano..."

2020-02-22T22:38:57Z

Created page with "'''Iterated amplification''' (also called '''iterated distillation and amplification''', and abbreviated '''IDA''') is the technical alignment agenda that Paul Christiano..."

New page

'''Iterated amplification''' (also called '''iterated distillation and amplification''', and abbreviated '''IDA''') is the technical alignment agenda that [[Paul Christiano]] works on.

==See also==

* [[My understanding of how IDA works]]

← Older revision		Revision as of 03:58, 26 April 2020
Line 37:		Line 37:

	[[Category:Iterated amplification]]		[[Category:Iterated amplification]]
		+	[[Category:AI safety]]

@@ Line 6: / Line 6: @@
 * [[adequate oversight]]
 * [[overseer]]
-* [[bandwidth of the overseer]]
+* [[bandwidth of the overseer]], [[high bandwidth oversight]], [[low bandwidth oversight]]
 * [[reward engineering]]
 * [[HCH]], [[Strong HCH]], [[Weak HCH]], [[Humans consulting HCH]]

@@ Line 30: / Line 30: @@
 * [[universality]]
 * [[narrow value learning]] vs [[ambitious value learning]]
 ==See also==

← Older revision		Revision as of 22:47, 24 February 2020
Line 1:		Line 1:
	'''Iterated amplification''' (also called '''iterated distillation and amplification''', and abbreviated '''IDA''') is the technical alignment agenda that [[Paul Christiano]] works on.		'''Iterated amplification''' (also called '''iterated distillation and amplification''', and abbreviated '''IDA''') is the technical alignment agenda that [[Paul Christiano]] works on.
		+
		+	Terminology (not necessarily about IDA, but these are some terms frequently used by Paul):
		+
		+	* [[informed oversight]]
		+	* [[adequate oversight]]
		+	* [[overseer]]
		+	* [[bandwidth of the overseer]]
		+	* [[reward engineering]]
		+	* [[HCH]], [[Strong HCH]], [[Weak HCH]], [[Humans consulting HCH]]
		+	* [[amplification]]
		+	* [[capability amplification]]
		+	* [[distillation]]
		+	* [[factored cognition]], [[factored evaluation]], [[factored generation]]
		+	* [[corrigibility]]
		+	* [[benign]]
		+	* [[aligned]]
		+	* [[robustness]]
		+	* [[red teaming]]
		+	* [[ALBA]]
		+	* [[optimization daemon]]s
		+	* [[act-based agent]] vs [[goal-directed agent]]
		+	* [[approval-directed agent]]
		+	* [[steering problem]]
		+	* [[prosaic AI]]
		+	* [[bootstrapping]]
		+	* [[catastrophe]]
		+	* [[reliability amplification]]
		+	* [[security amplification]]
		+	* [[universality]]
		+	* [[narrow value learning]] vs [[ambitious value learning]]

	==See also==		==See also==

← Older revision		Revision as of 22:41, 22 February 2020
Line 4:		Line 4:

	* [[My understanding of how IDA works]]		* [[My understanding of how IDA works]]
		+
		+	[[Category:Iterated amplification]]