Search IconIcon to open search

Embedded AI

Last updated Jul 2, 2021 Edit Source

More in post on AI systems

# Dreyfus

Dreyfus believed that, for any AI system to achieve any sort of general intelligence, it must also exhibit Dasein (being in the world). Thus, “a successful Heideggerian AI would need a perfect model of the human body – and by implication, that Dasein must be expressed as a human being, organically as well as existentially”.

Dreyfus’s critique of the frame approach → descriptions of typical situations like going to a birthday party

# Approaches to embedded AI

# Rodney Brook’s behaviourist approach

It turns out to be very difficult to reproduce in an internal representation for a computer the necessary richness of environment that would give rise to interesting behaviour by a highly adaptive robot

This is generally avoided by human beings because their model of the world is the world itself

Solution is to build a mobile robot that uses the world itself as its own representation (referring to its sensors rather than to an internal world model)


# Phil Agre’s pragmatist model

Use of deictic representations: instead of representing a particular object in the world, we represent a role that an object might play in a certain time-extended pattern of interaction between an agent and its environment

e.g. when a virtual ice cube defined by its functions is close to the virtual player, a rule dictates a response, e.g. kick it


# Merleau-Ponty’s discriminatory model

# Walter Freeman’s neurodynamic model