Computer Science
The study of computation, algorithms, and artificial intelligence illuminates the nature and limits of mechanism—and suggests that mind transcends machine.
Artificial Intelligence and Understanding
AI research illuminates both the power and limits of computational approaches to intelligence. Modern AI achieves remarkable feats—but does it understand? Does it think? Does it experience?
The Chinese Room argument suggests that syntax manipulation doesn't produce semantics. A computer manipulating symbols according to rules doesn't understand what the symbols mean. Understanding requires something more than computation—perhaps consciousness, perhaps intentionality.
- The Chinese Room: Searle's famous argument: a person following rules to manipulate Chinese symbols doesn't understand Chinese. Neither does a computer. Syntax doesn't produce semantics.
- The Symbol Grounding Problem: How do symbols acquire meaning? Computers manipulate symbols without understanding them. The symbols are meaningless to the machine—meaning exists only for the programmer and user.
- The Consciousness Gap: No AI system has subjective experience—the hard problem remains unsolved. We can simulate behavior without creating experience. There is no 'what it is like' to be a computer.
- Intentionality: Mental states are 'about' things—they have intentionality. Thoughts refer to objects. Computers don't have genuine intentionality; their 'aboutness' is derived from human minds.
Design, Information, and Intelligence
Software engineering reveals principles that apply to biological systems. Functional software requires intelligent design—random processes don't produce working code.
DNA is often compared to software—a digital code that specifies the construction of organisms. If software requires a programmer, what does this suggest about the genetic code? The analogy between biological and computational information points toward intelligent origin.
- Specified Complexity: Functional software exhibits specified complexity—it is both improbable and matches an independent specification. Random processes don't produce working programs.
- Information Origin: All known sources of specified information are intelligent agents. We never observe random processes generating functional code. This is a uniform experience.
- DNA as Code: Genetic information resembles software—a quaternary digital code with error correction, compression, and modular design. Bill Gates said DNA is like a software program, only much more complex.
- The Programmer Inference: If we found a computer program on Mars, we would infer intelligence. DNA is a more sophisticated program. The inference to intelligence is the same.