Van der Maas, How much intelligence is there in artificial intelligence? A 2020 update
6 important questions on Van der Maas, How much intelligence is there in artificial intelligence? A 2020 update
How does Dennett's cognitive wheel illustrate that ai research may be irrelivant for human intelligence research?
- Dennett's cognitive wheel states this: The wheel is a great intervention which facilitates movement with incredible efficiency. It does not however, learn us anything about how movement in nature occurs.
- AI may be building intelligent machines, but that machine intelligence is not similar to human intelligence or cognition.
- AI has incredible performance on various tasks but does not learn us anything about how humans perform these tasks since their capabilities/strategies are very different.
What is a second argument why ai research may be irrelevant for human intelligence research?
- AI can only inform us about cognition and not intelligence, as the latter primarily concerns individual differences in cognitive functioning, and not cognitive functioning itself.
Why is deep blue, the chess computer that beat kasparov, a prime example of a cognitive wheel?
- Deep Blue mainly tought is how chess players do not think.
- It applies tree search algorithms down all possible moves and evaluates all positions n moves deep. It picks the move that results in the highest chance of a favorable position.
- Humans are unable to systematically explore all possible combinations of moves in a certain position.
- Higher grades + faster learning
- Never study anything twice
- 100% sure, 100% understanding
How did alpha zero chess improve on deep blue's approach?
- Instead of the brute force computing by which deep blue chooses a move, alpha zero chess consists of a deep learning network that learns how to play chess through a reinforcement learning system
What are artifical neural networks and what are their origins?
- ANNs are composed of a collection of largely identical processing “units” structured in a hierarchical fashion, similar to how biological neurons make up biological neural pathways.
- Psychology and neuroscience had, and still have, a strong influence on the development of ANNs.
- The development of ANNs in the 1950s was originally inspired by the structure and function of biological neurons and networks (McCulloch & Pitts, 1943).
- The goal was to model cognition using models that stayed true to what, at the time, was known about brain processing. This approach is called “connectionism”
Is AI intelligent according to the authors?
- Even though ai has superhuman performance on various tasks, the argument posed by searle still holds for modern ai. AI still lacks intelligence.
- This is illustrated with anti-computer positions. human chess players quickly see the point of the position, while AI systems are only able to evaluate the position after considering long sequences of moves.
- It wouldn't necessarily take a lot of training to have computer chess players recognize these position's faster, but according to Searle, Dreyfus and others, the problem lies deeper. AI systems lack semantics, feelings, a body and goals.
The question on the page originate from the summary of the following study material:
- A unique study and practice tool
- Never study anything twice again
- Get the grades you hope for
- 100% sure, 100% understanding