The Promise & Peril Of The Coming Technological Singularity
How accelerating change will lead to sentient machines & why it’s creating existential uncertainty about the future of mankind.
The singularity is the digital nirvana of the 21st century: a moment in time when machines become self-aware, using human-level capabilities at machine speeds to rapidly bootstrap themselves to godlike levels of intelligence — and what happens after that is anyone’s guess.
How you feel about the singularity depends on your perspective — almost like a Rorschach test for futurists. Optimists like Ray Kurzweil, Mark Zuckerberg, and Larry Page see it as the birth of a new utopian future for mankind, while others such as Nick Bostrom, Bill Gates & Elon Musk are deeply concerned.
Despite the uncertainties, science & technology continue their march forward — with numerous Artificial General Intelligence (AGI) projects and efforts to reverse engineer the human mind pushing us closer to that day of reckoning.
As Marshall Brain has written, “we’re creating a second intelligent species to compete with ourselves” — and while we know little about the potential thoughts, motives, or perhaps feelings that it will have when it arrives, we at least have a notion of the timeline: futurist Ray Kurzweil predicts it will happen by the year 2045.
The concept of the singularity originates with John von Neumann back in the ‘50s, but it remained obscure until science-fiction author Vernor Vinge popularized it in his 1993 essay, The Coming Technological Singularity, and subsequently in Ray Kurzweil’s The Age Of Spiritual Machines in 1999.
While predictions for the singularity vary, the premise arises from a common observation: the rate of technological advance is increasing exponentially — and at some point humans will simply be unable to keep up. The image above, for example, shows Moore’s Law creating an exponential increase in computing speed in a standard $1,000 desktop PC, which is plotted against milestones representing the computing capacity of animals & people.
While accelerating change in today’s world is created by people using machine-assisted design and online sharing of information, tomorrow it will be entirely the product of recursive self-improvement by machines, at which point humans will become irrelevant to progress — and pessimists worry that mankind will be irrelevant to the planet as well.
The Birth of Machine Superintelligence
Futurist Ray Kurzweil predicts machines will reach human level intelligence by the year 2030 — and after that point, computing power will continue to increase, surpassing the power of all human minds combined by around 2060.
Kurzweil’s analysis doesn’t address the complexities of developing software capable of mimicking the human mind — but he believes that advances in neuroscience will facilitate the development of true artificial intelligence, and projects like Carnegie Mellon’s IARPA-funded brain research project appear to support his prediction.
While computing capacity is not synonymous with sentience, Kurzweil believes that combined progress in computing & artificial intelligence will inevitably lead a self-aware, self-improving machine that Kurzweil predicts will lead to a Technological Singularity, “a hypothetical future point in time at which technological growth becomes uncontrollable and irreversible, resulting in unforeseeable changes to human civilization.”
As mentioned above, advances in artificial intelligence and the prospect of the singularity have been polarizing in the futurism community. The graph below, from a Vanity Fair article on the dangers of AI, shows a diverse spectrum of opinions on the topic held by several popular futurists.
Perhaps the most optimistic of these thinkers is Ray Kurzweil, who’s said, “Biological evolution is too slow for the human species. Over the next few decades, it’s going to be left in the dust. By the 2030, the non-biological portion of our intelligence — our technology — will predominate…”
In Kurzweil’s view, the singularity isn’t an event where machines rise to dominance over mankind, but rather the merger of man and machine, blurring the distinction between natural and artificial, and in the process re-defining life as we know it.
Kurzweil believes that concepts like mind uploading, molecular nanotechnology & cybernetic implants open the door to virtual immortality for humans, along with increased human intelligence, the end of disease, and a panacea of incredible new possibilities we haven’t even begun to explore.
Obviously not all futurists are not so optimistic. The late physicist Stephen Hawking held a very pessimistic view of AI, saying, “The development of full artificial intelligence could spell the end of the human race”, and he wasn’t alone in his concerns.
Elon Musk has called AI an “existential threat to humanity”, and Nick Bostrom has said, “Our approach to existential risks cannot be one of trial-and-error. There is no opportunity to learn from errors.”
A more moderate approach comes from Bill Gates, calling AI “both promising and dangerous”, but still advocating open, managed research. “The world hasn’t had that many technologies that are both promising and dangerous, he has said, “we had nuclear energy and nuclear weapons…the place that I think this is most concerning is in weapon systems.”
When Will The Singularity Happen?
Ray Kurzweil, being perhaps the best-known prognosticator of the singularity, has said that computers will match human levels of intelligence by the year 2029 — with the singularity itself anticipated to happen before the year 2045.
Other futurists offer slightly different timelines, with SoftBank CEO Masayoshi Son saying it will happen in 2047, and Ben Goertzel, an advisor to Singularity University, claiming that various social & scientific factors could put the date anywhere from 2020 to 2100.
Futurism.com recently polled several computing & AI experts in an article entitled, Separating Science Fact From Science Hype: How Far off Is the Singularity?, and the consensus was generally consistent with Kurzweil’s projections, with a timeframe between 2045 and 2050.
It’s worth noting that Kurzweil himself padded his view of the timeline for the singularity with the premise that even if human-level machine intelligence doesn’t happen immediately, the Law Of Accelerating Returns means that only a short delay will exist before the singularity occurs — a result of the exponential rate of increase Kurzweil is fond of citing.
Exponential growth in computing capacity over the last century points to an inexorable conclusion — that computers will match and eventually exceed the computing capacity of the human mind.
At some point, this increased computing power combined with advances in artificial intelligence is predicted to lead to the creation of a self-aware machine intelligence, which will be capable of recursive self-improvement to bootstrap itself to levels of intelligence beyond the scope of human imagination.
The result will be one of the most important events in the history of the human race: the singularity will be the birth of a second intelligent species on Earth, capable of competing with mankind and driven by goals, desires & motivations that cannot be predicted.
Various futurists hold an array of opinions about the meaning & impact of the singularity — with some viewing it as a desired outcome capable of elevating mankind to godlike stature, and others anticipating the complete displacement of human beings by a new, more advanced form of intelligence.
Currently not enough is understood about the potential impact of the singularity to draw a definite conclusion, but perhaps the best way to regard the it is like any powerful new technology: a pandora’s box of mixed blessings, fraught with an uncertain number of benefits & dangers.