4 - Intelligence without representation - Brooks (1991)
Dit paper bespreekt de mogelijkheid van intelligentie zonder representatie.
• Wat is precies het verschil tussen intelligentie met en intelligentie zonder representatie?
Kun je hier een voorbeeld van geven?
• Wat is volgens de auteur het voordeel van intelligente systemen zonder representatie?
• De auteur stelt een notie voor van intelligentie zonder representatie. Hoe verhoudt deze
notie van intelligentie zich tot de vier definities van AI uit het 1e paper dat jullie hebben
gelezen? Is deze notie compatibel met alle vier de definities, of niet?
- AI started as a field whose goal was to replicate human level intelligence in a machine
- approach to creating artificial intelligence:
- built up the capabilities of intelligent systems, having complete systems at each step of
the way and thus automatically ensure that the pieces and their interfaces are valid
- at each step we should build complete intelligent systems that we let loose in the real
world with real sensing and real action
→ conclusion: explicit representation and models of the world get in the way of simple level
intelligence, it’s better to use the world as its own model
→ hypothesis: representation is the wrong unit of abstraction in building the bulkiest parts of
intelligent systems
- in AI, abstraction is used to factor out all aspects of perception and motor skills
- by representing only the pertinent (relevant) facts explicitly, the semantics (explanations) of a
world were reduced to a simple closed system once again
- the only input to most AI programs is a restricted set of simple assertions deduced from the
real data by humans → recognition, spatial understanding, dealing with sensor noise & partial
models are all ignored
- blocks world: a simple and abstract world in AI, blocks (in different colors and shapes) are
placed on a surface, it is used to test AI-algorithms and understand them
- Merkwelt: a perceptual world; concept dat beschrijft hoe een creature in staat is om dingen te
bekijken, manipuleren en synthetiseren om vervolgens de betekenis te interpreteren van dingen
in de wereld om hem heen.
→ each animal and robot species has their own different Merkwelt
→ the Merkwelt we humans provide our programs is based on our own introspection
→ objections:
1. danger that reasoning strategies developed for the human-assumed Merkwelt may not be
valid when real sensors and perception processing is used
Dit paper bespreekt de mogelijkheid van intelligentie zonder representatie.
• Wat is precies het verschil tussen intelligentie met en intelligentie zonder representatie?
Kun je hier een voorbeeld van geven?
• Wat is volgens de auteur het voordeel van intelligente systemen zonder representatie?
• De auteur stelt een notie voor van intelligentie zonder representatie. Hoe verhoudt deze
notie van intelligentie zich tot de vier definities van AI uit het 1e paper dat jullie hebben
gelezen? Is deze notie compatibel met alle vier de definities, of niet?
- AI started as a field whose goal was to replicate human level intelligence in a machine
- approach to creating artificial intelligence:
- built up the capabilities of intelligent systems, having complete systems at each step of
the way and thus automatically ensure that the pieces and their interfaces are valid
- at each step we should build complete intelligent systems that we let loose in the real
world with real sensing and real action
→ conclusion: explicit representation and models of the world get in the way of simple level
intelligence, it’s better to use the world as its own model
→ hypothesis: representation is the wrong unit of abstraction in building the bulkiest parts of
intelligent systems
- in AI, abstraction is used to factor out all aspects of perception and motor skills
- by representing only the pertinent (relevant) facts explicitly, the semantics (explanations) of a
world were reduced to a simple closed system once again
- the only input to most AI programs is a restricted set of simple assertions deduced from the
real data by humans → recognition, spatial understanding, dealing with sensor noise & partial
models are all ignored
- blocks world: a simple and abstract world in AI, blocks (in different colors and shapes) are
placed on a surface, it is used to test AI-algorithms and understand them
- Merkwelt: a perceptual world; concept dat beschrijft hoe een creature in staat is om dingen te
bekijken, manipuleren en synthetiseren om vervolgens de betekenis te interpreteren van dingen
in de wereld om hem heen.
→ each animal and robot species has their own different Merkwelt
→ the Merkwelt we humans provide our programs is based on our own introspection
→ objections:
1. danger that reasoning strategies developed for the human-assumed Merkwelt may not be
valid when real sensors and perception processing is used
