Skip to main content
Artificial Intelligence

Return to Answer

I tried to describe the approaches more in detail.
Source Link
nbro
nbro
  • 41.4k
  • 12
  • 114
  • 205

The paper Artificial General Intelligence: Concept, State of the Art, and Future Prospects (2014), by Ben Goertzel (one of the people that are really still very interested in AGI), surveys the field of artificial general intelligence (AGI), its progress, approaches, mathematical formalisms, engineering, and biology-inspired perspectives, and metrics for assessing AGI.

Just to give a little bit more context and whet your appetite, let me briefly describe the different approaches to AGI (section 3, p. 14).

  • symbolic approach (which is based on the Physical Symbol System Hypothesis; examples of this approach are ACT-R or SOAR),

  • emergentist approach (aka sub-symbolic, i.e. the use of neural networks, and similar sub-symbolic models, from which abstract symbolic processing/reasoning can or is expected to emerge; so examples of this approach is deep learning, computational neuroscience, and artificial life),

  • hybrid approach (a combination of the symbolic and sub-symbolic approaches; examples of this approach are CLARION and CogPrime), and

  • universalist approach (examples of this approach are the AIXI and Gödel machine).

The paper Artificial General Intelligence: Concept, State of the Art, and Future Prospects (2014), by Ben Goertzel (one of the people that are really still very interested in AGI), surveys the field of artificial general intelligence (AGI), its progress, approaches, mathematical formalisms, engineering, and biology-inspired perspectives, and metrics for assessing AGI.

Just to give a little bit more context and whet your appetite, let me briefly describe the different approaches to AGI (section 3, p. 14).

  • symbolic approach (which is based on the Physical Symbol System Hypothesis; examples of this approach are ACT-R or SOAR),

  • emergentist approach (aka sub-symbolic, i.e. the use of neural networks, and similar sub-symbolic models, from which abstract symbolic processing/reasoning can emerge; so examples of this approach is deep learning, computational neuroscience, and artificial life),

  • hybrid approach (a combination of the symbolic and sub-symbolic approaches; examples of this approach are CLARION and CogPrime), and

  • universalist approach (examples of this approach are the AIXI and Gödel machine).

The paper Artificial General Intelligence: Concept, State of the Art, and Future Prospects (2014), by Ben Goertzel (one of the people that are really still very interested in AGI), surveys the field of artificial general intelligence (AGI), its progress, approaches, mathematical formalisms, engineering, and biology-inspired perspectives, and metrics for assessing AGI.

Just to give a little bit more context and whet your appetite, let me briefly describe the different approaches to AGI (section 3, p. 14).

  • symbolic approach (which is based on the Physical Symbol System Hypothesis; examples of this approach are ACT-R or SOAR),

  • emergentist approach (aka sub-symbolic, i.e. the use of neural networks, and similar sub-symbolic models, from which abstract symbolic processing/reasoning can or is expected to emerge; so examples of this approach is deep learning, computational neuroscience, and artificial life),

  • hybrid approach (a combination of the symbolic and sub-symbolic approaches; examples of this approach are CLARION and CogPrime), and

  • universalist approach (examples of this approach are the AIXI and Gödel machine).

I tried to describe the approaches more in detail.
Source Link
nbro
nbro
  • 41.4k
  • 12
  • 114
  • 205

The paper Artificial General Intelligence: Concept, State of the Art, and Future Prospects (2014), by Ben Goertzel (one of the people that are really still very interested in AGI), surveys the field of artificial general intelligence and(AGI), its progress. It reviews, approaches, mathematical formalisms, engineering, and biology-inspired perspectives. There are several approaches to AGI, including symbolic, emergentist, hybrid and universalist ones. The paper also discusses metrics for assessing AGI.

Just to give a little bit more context and whet your appetite, let me briefly describe the different approaches to AGI (section 3, p. 14).

  • symbolic approach (which is based on the Physical Symbol System Hypothesis; examples of this approach are ACT-R or SOAR),

  • emergentist approach (aka sub-symbolic, i.e. the use of neural networks, and similar sub-symbolic models, from which abstract symbolic processing/reasoning can emerge; so examples of this approach is deep learning, computational neuroscience, and artificial life),

  • hybrid approach (a combination of the symbolic and sub-symbolic approaches; examples of this approach are CLARION and CogPrime), and

  • universalist approach (examples of this approach are the AIXI and Gödel machine).

The paper Artificial General Intelligence: Concept, State of the Art, and Future Prospects (2014), by Ben Goertzel, surveys the field of artificial general intelligence and its progress. It reviews mathematical formalisms, engineering and biology-inspired perspectives. There are several approaches to AGI, including symbolic, emergentist, hybrid and universalist ones. The paper also discusses metrics for assessing AGI.

The paper Artificial General Intelligence: Concept, State of the Art, and Future Prospects (2014), by Ben Goertzel (one of the people that are really still very interested in AGI), surveys the field of artificial general intelligence (AGI), its progress, approaches, mathematical formalisms, engineering, and biology-inspired perspectives, and metrics for assessing AGI.

Just to give a little bit more context and whet your appetite, let me briefly describe the different approaches to AGI (section 3, p. 14).

  • symbolic approach (which is based on the Physical Symbol System Hypothesis; examples of this approach are ACT-R or SOAR),

  • emergentist approach (aka sub-symbolic, i.e. the use of neural networks, and similar sub-symbolic models, from which abstract symbolic processing/reasoning can emerge; so examples of this approach is deep learning, computational neuroscience, and artificial life),

  • hybrid approach (a combination of the symbolic and sub-symbolic approaches; examples of this approach are CLARION and CogPrime), and

  • universalist approach (examples of this approach are the AIXI and Gödel machine).

deleted 12 characters in body
Source Link
nbro
nbro
  • 41.4k
  • 12
  • 114
  • 205

The paper Artificial General Intelligence: Concept, State of the Art, and Future Prospects (2014), by Ben Goertzel, surveys the diverse communityfield of artificial general intelligence and its progress. It reviews mathematical formalisms, engineering and biology-inspired perspectives. There are several approaches to AGI, including symbolic, emergentist, hybrid and universalist ones. The paper also discusses metrics for assessing AGI.

The paper Artificial General Intelligence: Concept, State of the Art, and Future Prospects (2014), by Ben Goertzel, surveys the diverse community of artificial general intelligence and its progress. It reviews mathematical formalisms, engineering and biology-inspired perspectives. There are several approaches to AGI, including symbolic, emergentist, hybrid and universalist ones. The paper also discusses metrics for assessing AGI.

The paper Artificial General Intelligence: Concept, State of the Art, and Future Prospects (2014), by Ben Goertzel, surveys the field of artificial general intelligence and its progress. It reviews mathematical formalisms, engineering and biology-inspired perspectives. There are several approaches to AGI, including symbolic, emergentist, hybrid and universalist ones. The paper also discusses metrics for assessing AGI.

Source Link
nbro
nbro
  • 41.4k
  • 12
  • 114
  • 205