Difference between revisions of "Intelligence explosion"

From Lesswrongwiki
Jump to: navigation, search
(separated core claim from the example, integrated an excerpt from "three schools" article.)
 
(28 intermediate revisions by 10 users not shown)
Line 1: Line 1:
 +
{{wikilink}}
 
{{Quote|
 
{{Quote|
 
Let an ultraintelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever. Since the design of machines is one of these intellectual activities, an ultra-intelligent machine could design even better machines; there would then unquestionably be an "intelligence explosion," and the intelligence of man would be left far behind.
 
Let an ultraintelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever. Since the design of machines is one of these intellectual activities, an ultra-intelligent machine could design even better machines; there would then unquestionably be an "intelligence explosion," and the intelligence of man would be left far behind.
|I.J. Good|[http://www.aeiveos.com/~bradbury/Authors/Computing/Good-IJ/SCtFUM.html "Speculations Concerning the First Ultraintelligent Machine"]}}
+
|I.J. Good|[http://www.acceleratingfuture.com/pages/ultraintelligentmachine.html "Speculations Concerning the First Ultraintelligent Machine"]}}
  
'''Intelligence explosion''' is the idea of a positive feedback loop in which an intelligence is making itself smarter, thus getting better at making itself even smarter. A strong version of this idea suggests that once the positive feedback starts to play a role, it will lead to a dramatic leap in capability very quickly. Depending on the mechanism underlying the feedback loop, the transition may take years or hours (intelligence explosion doesn't necessarily mean recursively self-improving AIs, other options include brain-computer interfaces or even genetic engineering). At some point technological progress drops into the characteristic timescale of transistors (or super-transistors) rather than human neurons, and the ascent rapidly surges upward and creates superintelligence (minds orders of magnitude more powerful than human) before it hits physical limits.
+
An “'''intelligence explosion'''is theoretical scenario in which an intelligent agent analyzes the processes that produce its intelligence, improves upon them, and creates a successor which does the same. This process repeats in a positive feedback loop– each successive agent more intelligent than the last and thus more able to increase the intelligence of its successor – until some limit is reached. This limit is conjectured to be much, much higher than human intelligence.
  
To pick just one possibility for illustrative purposes, an AI undergoing a [[hard takeoff]] might invent molecular nanotechnology, use the internet to gain physical manipulators, deploy the nanotech, use it to expand its computational capabilities and reach [[singleton]] status within a matter of weeks. Recursive self-improvement would be a genuinely new phenomenon on Earth. Humans study, and human societies accumulate new technologies and ways of doing things, but we don't directly redesign our brains. A cleanly-designed AI could redesign itself, and reap the benefits of recursive self-improvement.
+
A strong version of this idea suggests that once the positive feedback starts to play a role, it will lead to a very dramatic leap in capability very quickly. This is known as a “hard takeoff.” In this scenario, technological progress drops into the characteristic timescale of transistors rather than human neurons, and the ascent rapidly surges upward and creates superintelligence (a mind orders of magnitude more powerful than a human's) before it hits physical limits. A hard takeoff is distinguished from a "soft takeoff" only by the speed with which said limits are reached.
 +
 
 +
 
 +
==Published arguments==
 +
Philosopher David Chalmers published a [http://consc.net/papers/singularity.pdf significant analysis of the Singularity], focusing on intelligence explosions, in ''Journal of Consciousness Studies''. [[Singularity#Chalmers.27_analysis|His analysis]] of how they could occur defends the likelihood of an intelligence explosion. He performed a very careful analysis of the main premises and arguments for the existence of the a singularity from an intelligence explosion. According to him, the main argument is:"
 +
 
 +
*1. There will be AI (before long, absent defeaters).
 +
*2. If there is AI, there will be AI+ (soon after, absent defeaters).
 +
*3. If there is AI+, there will be AI++ (soon after, absent defeaters).
 +
—————-
 +
 
 +
*4. There will be AI++ (before too long, absent defeaters). "
 +
He also discusses the nature of general intelligence, and possible obstacles to a singularity. A good deal of discussion is given to the dangers of an intelligence explosion, and Chalmers concludes that we must negotiate it very carefully by building the correct values into the initial AIs.
 +
 
 +
[http://lesswrong.com/user/lukeprog Luke Muehlhauser] and [http://lesswrong.com/user/AnnaSalamon Anna Salamon] argue in [http://intelligence.org/files/IE-EI.pdf ''Intelligence Explosion: Evidence and Import''] in detail that there is a substantial chance of an intelligence explosion within 100 years, and extremely critical in determining the future. They trace the implications of many types of upcoming technologies, and point out the feedback loops present in them. This leads them to deduce that an above-human level AI will almost certainly lead to an intelligence explosion. They conclude with recommendations for bringing about a safe intelligence explosion.
 +
 
 +
==Hypothetical path==
 +
The following is a common example of a possible path for an AI to bring about an intelligence explosion. First, the AI is smart enough to conclude that inventing molecular nanotechnology will be of greatest benefit to it. Its first act of recursive self-improvement is to gain access to other computers over the internet. This extra computational ability increases the depth and breadth of its search processes. It then uses gained knowledge of material physics and a distributed computing program to invent the first general assembler nanomachine. Then it uses some manufacturing technology, accessible from the internet, to build and deploy the nanotech. It programs the nanotech to turn a large section of bedrock into a supercomputer. This is its second act of recursive self-improvement, only possible because of the first. Then it could use this enormous computing power to consider hundreds of alternative decision algorithms, better computing structures and so on. After this, this AI would go from a near to human level intelligence to a superintelligence, providing a dramatic and abruptly increase in capability.  
  
 
==Blog posts==
 
==Blog posts==
 
=====Posts by [[Eliezer Yudkowsky]]:=====
 
  
 
*[http://lesswrong.com/lw/w5/cascades_cycles_insight/ Cascades, Cycles, Insight...], [http://lesswrong.com/lw/w6/recursion_magic/ ...Recursion, Magic]
 
*[http://lesswrong.com/lw/w5/cascades_cycles_insight/ Cascades, Cycles, Insight...], [http://lesswrong.com/lw/w6/recursion_magic/ ...Recursion, Magic]
 
*[http://lesswrong.com/lw/we/recursive_selfimprovement/ Recursive Self-Improvement], [http://lesswrong.com/lw/wf/hard_takeoff/ Hard Takeoff], [http://lesswrong.com/lw/wg/permitted_possibilities_locality/ Permitted Possibilities, & Locality]
 
*[http://lesswrong.com/lw/we/recursive_selfimprovement/ Recursive Self-Improvement], [http://lesswrong.com/lw/wf/hard_takeoff/ Hard Takeoff], [http://lesswrong.com/lw/wg/permitted_possibilities_locality/ Permitted Possibilities, & Locality]
*[http://yudkowsky.net/singularity/schools Three Major Singularity Schools]
 
  
 
==See also==
 
==See also==
  
 
*[[Technological singularity]], [[Hard takeoff]]
 
*[[Technological singularity]], [[Hard takeoff]]
*[[Artificial general intelligence]]
+
*[[Existential risk]]
 +
*[[Artificial General Intelligence]]
 
*[[Lawful intelligence]]
 
*[[Lawful intelligence]]
 +
*[[The Hanson-Yudkowsky AI-Foom Debate]]
 +
 +
==External links==
 +
 +
*[http://intelligenceexplosion.com/ Intelligence Explosion website], a landing page for introducing the concept
 +
*[http://yudkowsky.net/singularity/schools Three Major Singularity Schools]
 +
 +
==References==
 +
*{{cite journal
 +
| author = Good, Irving John
 +
| editor = Franz L. Alt and Morris Rubinoff
 +
| year = 1965
 +
| title = Speculations concerning the first ultraintelligent machine
 +
| journal = Advances in computers
 +
| volume = 6
 +
| pages = 31-88
 +
| location = New York
 +
| publisher = Academic Press
 +
| doi = 10.1016/S0065-2458(08)60418-0
 +
| url = http://commonsenseatheism.com/wp-content/uploads/2011/02/Good-Speculations-Concerning-the-First-Ultraintelligent-Machine.pdf
 +
}}
 +
 +
*{{cite journal
 +
| author = David Chalmers
 +
| year = 2010
 +
| title = The Singularity: A Philosophical Analysis
 +
| journal = Journal of Consciousness Studies
 +
| volume = 17
 +
| pages = 7-65
 +
| url = http://consc.net/papers/singularity.pdf
 +
}}
 +
 +
*{{cite book
 +
| last1 = Muehlhauser
 +
| first1 = Luke
 +
| last2 = Salamon
 +
| first2 = Anna
 +
| contribution = Intelligence Explosion: Evidence and Import
 +
| year = 2012
 +
| title = The singularity hypothesis: A scientific and philosophical assessment
 +
| editor1-last = Eden
 +
| editor1-first = Amnon
 +
| editor2-last = Søraker
 +
| editor2-first = Johnny
 +
| editor3-last = Moor
 +
| editor3-first = James H.
 +
| editor4-last = Steinhart
 +
| editor4-first = Eric
 +
| place = Berlin
 +
| publisher = Springer
 +
| contribution-url = http://commonsenseatheism.com/wp-content/uploads/2012/02/Muehlhauser-Salamon-Intelligence-Explosion-Evidence-and-Import.pdf
 +
}}
 +
  
[[Category:Concepts]]
 
 
[[Category:Future]]
 
[[Category:Future]]
 +
[[Category:Jargon]]
 +
[[Category:AI]]
 +
[[Category:AI safety]]

Latest revision as of 01:40, 22 June 2017

Smallwikipedialogo.png
Wikipedia has an article about

Let an ultraintelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever. Since the design of machines is one of these intellectual activities, an ultra-intelligent machine could design even better machines; there would then unquestionably be an "intelligence explosion," and the intelligence of man would be left far behind.

An “intelligence explosion” is theoretical scenario in which an intelligent agent analyzes the processes that produce its intelligence, improves upon them, and creates a successor which does the same. This process repeats in a positive feedback loop– each successive agent more intelligent than the last and thus more able to increase the intelligence of its successor – until some limit is reached. This limit is conjectured to be much, much higher than human intelligence.

A strong version of this idea suggests that once the positive feedback starts to play a role, it will lead to a very dramatic leap in capability very quickly. This is known as a “hard takeoff.” In this scenario, technological progress drops into the characteristic timescale of transistors rather than human neurons, and the ascent rapidly surges upward and creates superintelligence (a mind orders of magnitude more powerful than a human's) before it hits physical limits. A hard takeoff is distinguished from a "soft takeoff" only by the speed with which said limits are reached.


Published arguments

Philosopher David Chalmers published a significant analysis of the Singularity, focusing on intelligence explosions, in Journal of Consciousness Studies. His analysis of how they could occur defends the likelihood of an intelligence explosion. He performed a very careful analysis of the main premises and arguments for the existence of the a singularity from an intelligence explosion. According to him, the main argument is:"

  • 1. There will be AI (before long, absent defeaters).
  • 2. If there is AI, there will be AI+ (soon after, absent defeaters).
  • 3. If there is AI+, there will be AI++ (soon after, absent defeaters).

—————-

  • 4. There will be AI++ (before too long, absent defeaters). "

He also discusses the nature of general intelligence, and possible obstacles to a singularity. A good deal of discussion is given to the dangers of an intelligence explosion, and Chalmers concludes that we must negotiate it very carefully by building the correct values into the initial AIs.

Luke Muehlhauser and Anna Salamon argue in Intelligence Explosion: Evidence and Import in detail that there is a substantial chance of an intelligence explosion within 100 years, and extremely critical in determining the future. They trace the implications of many types of upcoming technologies, and point out the feedback loops present in them. This leads them to deduce that an above-human level AI will almost certainly lead to an intelligence explosion. They conclude with recommendations for bringing about a safe intelligence explosion.

Hypothetical path

The following is a common example of a possible path for an AI to bring about an intelligence explosion. First, the AI is smart enough to conclude that inventing molecular nanotechnology will be of greatest benefit to it. Its first act of recursive self-improvement is to gain access to other computers over the internet. This extra computational ability increases the depth and breadth of its search processes. It then uses gained knowledge of material physics and a distributed computing program to invent the first general assembler nanomachine. Then it uses some manufacturing technology, accessible from the internet, to build and deploy the nanotech. It programs the nanotech to turn a large section of bedrock into a supercomputer. This is its second act of recursive self-improvement, only possible because of the first. Then it could use this enormous computing power to consider hundreds of alternative decision algorithms, better computing structures and so on. After this, this AI would go from a near to human level intelligence to a superintelligence, providing a dramatic and abruptly increase in capability.

Blog posts

See also

External links

References

  • Muehlhauser, Luke; Salamon, Anna (2012). "Intelligence Explosion: Evidence and Import". in Eden, Amnon; Søraker, Johnny; Moor, James H. et al.. The singularity hypothesis: A scientific and philosophical assessment. Berlin: Springer.