(FB 1545835406 Timestamp)
SHOW PREP: AI. RPL.
INITIAL:
JF’s THEORY. ai..modification…biology..
PART I.
- Fear of weapons – they will happen. As soon as the power to weight ratio is solved, Men are no longer much use.
- Fear of job displacement – we can compensate for most of it as long as we limit immigration.
- Limiting problems are cost/energy/chice-competition
PART II. What I am unsure about and what I am sure about.
UNSURE
- Unsure about limits of emergence as memory increases. Appears that symbolic recursive performs the same function at cost of state and precision, and increased demand for communication. Appears that consequent layers will produce same functionality without the information loss.
SURE
-
Continuous forecasting … of WHAT?
-
Go, chess, checkers, are ‘simple’ problems.
Speech recognition, driving, are simple problems.
Adaptation to continuous problems is HARD. 2. Decidability => Required => Intent
Consciousness = Social/Moral
Comprehensible – Human Action as Language (networks are not open to introspection).
Property as Moral Calculation
Hemispheric Competition 3. Problem of discovery is not intelligence but testing ideas.
Small number of fundamental innovations in the frame (darwin, einstein, menger, mathiness/probability/operations)
Small number of fundamental methods of lying. 4. Liability of human work doesn’t end. Nuclear devices, chemical weapons, hazardous chemicals etc.