L2: Connectionist vs. Symbolic AI, Darthmouth Workshop Flashcards
What is connectionist AI?
- Neural networks
- Inspired but not simulating human brain
- Neurons compute output independently
- Graphs of neurons
Describe McCulloch & Pitts binary threshold artificial neuron
Proposed in 1943
Binary inputs and outputs, weight = {-1,1}, always have a threshold t
If activation >= t then output 1 otherwise 0
Cannot represent XOR (only networks can)
Networks of binary threshold neurons can express any
sentence in propositional logic
What is symbolic AI?
o Knowledge as symbols
o Expressions as relations between symbols
o Knowledge base (conjunction of formulas where each formula asserts some knowledge)
o Rules to reason (derive new knowledge through manipulation) to create new expressions- we generally use first-order logic
Who grounded mathematics with formal logic?
Whitehead & Russell (1910-1913)- Principa Mathematics
Tried to ground foundation of mathematics using formal logic.
* Started with limited set of axioms and inference rules
* Derived everything else from there
What is the Logic Theorist?
First AI program of Newell, Simon, Shaw (1956), encoded axioma and rules of Principa Mathematics
Clear example of symbolic AI (could not prove all theorems in PM)
What were automatic computers initially used for?
First uses often numerical and for military purposes
* Code-breaking efforts
* Ballistic trajectories
* Computations for hydrogen bomb
* . . .
Describe the Dartmouth workshop
John McCarthy Marvin Minsky Nathaniel Rochester Claude Shannon
“We propose that a 2-month, 10-man study of artificial intelligence be carried out
during the summer of 1956 at Dartmouth College in Hanover, New Hampshire.
The study is to proceed on the basis of the conjecture that every aspect of learning or
any other feature of intelligence can in principle be so precisely described that a
machine can be made to simulate it.”
- Automatic Computers
- How Can a Computer be Programmed to Use a Language
- Neuron Nets
- Theory of the Size of a Calculation
- Self-Improvement
- Abstractions
- Randomness and Creativity
Not many documented technical outcomes