1993 essay “The Coming Technological Singularity” more than a decade earlier. Once the Singularity has been reached, Kurzweil says that machine intelligence will be infinitely the singularity is near when humans transcend biology pdf powerful than all human intelligence combined. Afterwards he predicts intelligence will radiate outward from the planet until it saturates the universe.
The Singularity is also the point at which machines intelligence and humans would merge. Exponential growth is deceptive, nearly flat at first until it hits what Kurzweil calls “the knee in the curve” then rises almost vertically. In fact Kurzweil believes evolutionary progress is super-exponential because more resources are deployed to the winning process. As an example of super-exponential growth Kurzweil cites the computer chip business.
Kurzweil says evolutionary progress looks smooth, but that really it is divided into paradigms, specific methods of solving problems. Each paradigm starts with slow growth, builds to rapid growth, and then levels off. As one paradigm levels off, pressure builds to find or develop a new paradigm. Kurzweil claims the whole world economy is in fact growing exponentially, although short term booms and busts tend to hide this trend.
A fundamental pillar of Kurzweil’s argument is that to get to the Singularity, computational capacity is as much of a bottleneck as other things like quality of algorithms and understanding of the human brain. Kurzweil feels the increase in the capacity of integrated circuits will probably slow by the year 2020. He feels confident that a new paradigm will debut at that point to carry on the exponential growth predicted by his law of accelerating returns. Kurzweil describes four paradigms of computing that came before integrated circuits: electromechanical, relay, vacuum tube, and transistors. Moore’s Law ends it will eventually rival the raw computing power of the human brain. 2045, the onset of the Singularity, he says same amount of money will buy one billion times more power than all human brains combined today. Kurzweil admits the exponential trend in increased computing power will hit a limit eventually, but he calculates that limit to be trillions of times beyond what is necessary for the Singularity.
Kurzweil notes that computational capacity alone will not create artificial intelligence. He asserts that the best way to build machine intelligence is to first understand human intelligence. The first step is to image the brain, to peer inside it. 2020s when it becomes possible to scan the brain from the inside using nanobots. Once the physical structure and connectivity information are known, Kurzweil says researchers will have to produce functional models of sub-cellular components and synapses all the way up to whole brain regions. The human brain is “a complex hierarchy of complex systems, but it does not represent a level of complexity beyond what we are already capable of handling”. Beyond reverse engineering the brain in order to understand and emulate it, Kurzweil introduces the idea of “uploading” a specific brain with every mental process intact, to be instantiated on a “suitably powerful computational substrate”.