Back in 1965, Gordon Moore, the co-founder of Intel, described an example of “Accelerating Change” known as Moore’s Law . In simple terms, this is “the observation that the number of transistors in a dense integrated circuit doubles about every two years”.
The trend Moore was observing was driven by improvements in the photo-lithography process used to manufacture semiconductors. If you double the lithography resolution, you exponentially scale (4x) the number of transistors a chip — which leads to an exponential increase in computing power as well.
In that context, Moore’s Law is nothing more than an observation based on a manufacturing process, and by correlation, it means that this trend shouldn’t have existed before photolithography, and it won’t exist in the future when transistor density reaches a certain maximum density.
Beyond Moore’s Law
Describing Moore’s Law as being the result of a semiconductor manufacturing process makes sense, but it may overly simplistic. The evidence for this comes from futurist Ray Kurzweil, who back-plotted over a century of trends in computing power and found that older non-transistor computers fit into the same exponential trend line as modern chips.
In The Age Of Spiritual Machines, Kurzweil reformulated Moore’s Law into The Law of Accelerating Returns, which is loosely based on Vernor Vinge’s hypothesis of exponentially accelerating change. In essence, these ideas suggest that the exponential change in computing power described by Moore’s Law is merely a small part of a much larger exponentially increasing trend.
Kurzweil’s connection between Moore’s Law and Vinge’s hypothesis is a paradigm shift, because it generalizes the idea of accelerating change— but separating this observed effect from the well-understood cause raises a lot of awkward questions in the process.
The biggest question is the most challenging: if Moore’s Law was driven by advances in photolithography for semiconductor circuits, when what exactly is driving it both before and after the semiconductor era Gordon Moore was observing?
Is Moore’s Law An Evolutionary Force?
Kurzweil’s explanation requires somewhat of a leap of faith: he speculated that “biological evolution leads to technology, which leads to computation, which leads to Moore’s law.”
In other words, the crux of his argument is that Moore’s Law is one observable component of a much larger trend towards accelerating intelligence, plotted in terms of computing power — and he’s arguing that it’s somehow driven by evolution itself.
If this is true, it also suggests that the exponential increase in “intelligence” (processing ability) is substrate-independent: meaning that it started with biological systems evolving slowly, and jumped to vacuum tubes, then transistors, and when that’s exhausted it will jump yet again — perhaps to Quantum Computers.
It’s an intriguing idea, but it may have some flaws: while the per-operation computing in a vacuum-tube may be faster than a human synapse, the overall computing power of the devices they reside in is much lower in capacity.
However, by plotting aeons of evolution onto graphs, Kurzweil’s graph shows an intriguing trend that may indicate more than simply wishful thinking.
Is Artificial Intelligence A New Form Of Life?
If Kurzweil is correct, then the biological evolution of life itself tends to follow a long-term pattern of exponential increase in complexity — and machine intelligence, followed by a “Technological Singularity” when machines leap beyond human levels of intelligence, and may be the next step in a natural evolutionary process.
Kurzweil himself suggested this, indicating that evolution may be changing to a new substrate for life much as chip-designers transitioned from vacuum-tubes to transistors decades ago. While this is somewhat reminiscent of an “intelligent design” argument, it also suggests that if evolution is the force behind this development, then like all evolution what’s being created is a new form of life.
This would be a true breakthrough, because for the first time on Earth we could point to a new form of silicon-based life built on fundamentally different principles as an alternative to the carbon wetware that has been evolving for hundreds of millions of years.
The End Of The Human Era?
Futurists such as Nick Bostrom, Stephen Hawking, Bill Gates and Elon Musk have all expressed concerns about AI being an existential risk that threatens the future of mankind, and the Singularity is one of the primary reasons for their concern.
Kurzweil, on the other hand, remains an optimist, and suggests that rather than seeing humanity displaced by intelligent machines, we will at some point merge with them into a new species capable of inhabiting carbon, silicon and everything in between.
Recent advances in artificial limbs, cochlear implants, and brain-computer interfaces to lend some support to Kurzweil’s view. Machine intelligence may appear as a discrete entity separate from its creators, but based on rapid advances in cybernetics, it’s not likely to remain that way for long.