Dynamics of Human and Machine Intelligence
Unlike machines, human cognitive capacities are biologically defined and stagnant at practicable timeframes.
Approximately 200,000 years ago Homo sapien emerged from the savannah. She was armed with an exceptionally developed neocortex, a prominent frontal lobe, and an advanced capacity for articulate speech and abstract reasoning. However, and unfortunately, since then she has stalled.
Fundamental human capacities have remained largely unchanged for those 200,000 years. Biologically, human capacities are essentially fixed. We can move, deduce, infer, see, hear, emote, and so on. A person can learn new skills but these are bound by our biologic capacities. We can use our deductive reasoning to wrap our heads around calculus and geometry but we can’t, unaided, see into the ultraviolet spectrum, or hear ultrasonic waves.
We can expand our capacities with technology but not without. Our inherent capacities are effectively stagnant.
On the other hand, technology’s capacities are dynamic. In fact they’re dynamic in only one direction: they grow. First computers beat us at chess, then Go, then Dota. Once won, barring some cataclysmic event, these capacities cannot be lost.
Implications on Superintelligence
We can depict the sets of machine capacities and human capacities in a Venn Diagram. On the left are all machine capacities and on the right, human capacities. Their intersection marks capacities that are potentially but not necessarily automatable. The circle on the right remains at a standstill in practicable timeframes. The circle on the left grows larger and creeps to the right. This growth is non-linear, it accelerates over time.
These dynamics imply that technology will catch up to us. They also imply we’ll be surpassed.
Why will machines eventually surpass us? First, the forward arrow of time is very long, that is we have a long time to work on this problem. Second, human intelligence is material meaning that it’s not magic; it’s in our skulls. It is therefore subject to study, reverse engineering, and the like. Third, we’ve already made progress in this direction. Fourth, our rate of progress has been swift. It isn’t linear, each advance enables the next. Fifth, we don’t need to fully understand how the brain works to emulate it. We made light bulbs before understanding electrons.
Probability favours the eventual emergence of superintelligent machines. Many in the field think it will happen within the century. But regarding automation, particularly in health care, superhuman is overkill.
Brandy was wrong. Almost does count. At least where redundancy is concerned.
Remember that a job consists of core and ancillary tasks. Core tasks are essential to the role, ancillary are not. A machine does not have to be capable of all tasks. If it is capable of almost all, specifically core tasks, the role is at risk of automation. [Footnote: Only “at risk” because an economic case for automation must still be had.]
Summary
Our capacities, both cognitive and motor, are defined by our biology. Our biology evolves in impracticable timeframes (hundreds of thousands and millions of years). Therefore, for practical purposes, human capacities are fixed and stagnant. In contrast machine capacities are dynamic. In fact, barring a cataclysmic event, they are dynamic in only one direction: they grow. It follows that machine capacities will incrementally creep into and ultimately supersede human capacities. This has already happened in several domains. Where machine capacities match or exceed human capacities, tasks are automatable. Where this occurs and an economic case exists for automation, we have technological unemployment.