Remember The Day Your Psych Professor Explained Affordance?
July 23, 2020
I read “A Concept in Psychology Is Helping AI to Better Navigate Our World.” That write up contained an explanation of the “Theory of Affordance,” whipped up by a psychology whiz named James J. Gibson.
To be candid, I have zero recollection of Psychology 101. Two tests. Bingo. Done! Because my class convened in a big lecture hall, I was able to do other work as the professor, who struck me as crazy as a Looney Toons cartoon character, talked and talked and talked.
So the Theory of Avoidance means:
when intelligent beings look at the world they perceive not simply objects and their relationships but also their possibilities. In other words, the chair “affords” the possibility of sitting. The water “affords” the possibility of swimming.
I think of this in terms of “context,” but I passed the course, ignoring the psychobabble. I might add that many of the students in that course developed and manifested a range of behaviors, often in lockstep with the wacko’s lectures. Go figure?
The article is about Google DeepMind. That unit of the world’s most loved enterprise is embracing the Affordance thing. Good move. AI remains useful in certain use cases. In others, AI is mostly handwaving and a joy ride for those whose parents wondered if their budding genius in math would ever get a job.
The write up states about the Googlers:
The researchers set up a simple virtual scenario. They placed a virtual agent in a 2D environment with a wall down the middle and had the agent explore its range of motion until it had learned what the environment would allow it to do—its affordances. The researchers then gave the agent a set of simple objectives to achieve through reinforcement learning, such as moving a certain amount to the right or to the left. They found that, compared with an agent that hadn’t learned the affordances, it avoided any moves that would cause it to get blocked by the wall partway through its motion, setting it up to achieve its goal more efficiently.
What’s next? More money, time, and development.
AI marches forward on a floor which affords the opportunity to step on a humanoid’s foot or possibly a crawling baby.
Stephen E Arnold, July 23, 2020
Comments
-
- Subscribe to Beyond Search
Feature archive
News archive
-
