ConceptAGIDoability

Related: Approach

5. AGI is not as big a development task as most think.

• a. DNA defines initial brain structure but not much DNA data is devoted to neocortex formation. Therefore, the brain (and an AGI) must be possible with repeating patterns of simpler neural circuits or simple rules which govern synapse creation.
• b. Brain capacity is bounded by neuron counts.
• c. Counts of sensory and motor nerves bound the incoming and outgoing data rates.
• d. These limitations set a maximum complexity for the brain and hence for an AGI.

6. AGI Requires Robotics. Without interaction with the real world, artificial intelligence will always be narrow. The real world is so variable and complex that simulators can speed development.
7. AGI can be created from existing hardware.

• a. Enough performance is available from today’s hardware. (SoftwareCompute)
• b. Some subset of human performance could qualify as AGI.

8. AGI will not be like human intelligence. Human intelligence develops within the context of human goals, emotions, and instincts, which would form a poor basis for AGI.

Some of the reasoning above is currently subject to dispute and may eventually prove to be in error. The development of the Brain Simulator can help settle philosophical disputes one way or another. At the same time, the structure of the Brain Simulator is flexible enough to adapt to new information as it becomes available.