Connectionist networks are formal systems

Arguments used against the formal character of symbol manipulators apply equally well to connectionist networks [CNs]. Functions computed on a CN can also be computed on a serial machine and CNs can implement classical serial processing.

Any function that can be computed on a connectionist network, can also be computed on a serial machine. In fact, most current connectionist networks are simulated on serial machines. Conversely, connectionist networks can be used to implement classical serial processing. Thus, arguments directed against the formal character of symbol manipulators apply equally well against connectionist networks.

Supported by "Mathematical insight is Non-Algorhythmic", Map 7, Box 23.

Note: the point the connectionist networks and symbol systems can simulate each other is widely accepted. Searle and Penrose use this point to show that the Chinese Room & Goedel arguments, respectively, apply to connectionist networks as well as to classical AI systems.
 
John Searle, 1990b & Roger Penrose, 1989.
RELATED ARTICLESExplain
Artificial Intelligence
Can computers think? [1]
Yes: connectionist networks can think [5a]
Connectionist networks are formal systems
The Chinese Gym Argument
Simulations of connectionist networks aren't duplications
John Searle
Roger Penrose
Connectionist networks can think without following rules
The Connectionist Biological Assumption
The Connectionist Dilemma
The Subsymbolic Paradigm
Connectionist computers lack commonsense
Connectionists fall into a computational mindset
One-layer perceptrons can’t compute certain functions
Graph of this discussion
Enter the title of your article


Enter a short (max 500 characters) summation of your article
Enter the main body of your article
Lock
+Comments (0)
+Citations (0)
+About
Enter comment

Select article text to quote
welcome text

First name   Last name 

Email

Skip