How near is the Singularity?

The Technological Singularity has been defined by many and in various ways, Vince defined it as “within thirty years, we will have the technological means to create superhuman intelligence. Shortly after, the human era will be ended. […] I think it’s fair to call this event a singularity” (Vinge, 1993)

The consensus of all the definitions is that at a moment in time, artificial intelligence will have progressed to the point of a greater-than-human intelligence.  The prediction is that when artificial intelligence (AI) exceeds human intelligence that computers will be dominate over humans.  To support this theory is the comparison of humans and apes put forward by Kurzweil.  (Kurzweil, 2005) Apes are physically stronger than humans yet its human’s greater intelligence that has allowed us dominate the earth.  This theory appears to be sound so it’s possible that once artificial intelligence exceeds humans that something similar could happen.  The tipping point when this occurs is referred to as the Singularity.

The use of the word Singularity to define this event is very interesting.  A singularity means a point where some property is infinite.  Singularity has its basis in maths, particularly algebra. A common example used is:

f (x)

We know that dividing by 0 (zero) is undefined (impossible to calculate) but x can be an ever increasing small number that approaches but never reaches 0 thus yielding larger values approaching infinity.  This is defined in maths as a singularity and the theory is that a singularity is not solvable.  Interesting then that we refer to an event we believe will definitely happen as a singularity!

Ray Kurzweil is a Director of Engineering at Google, technologist, futurist, has made many successful predictions about the future of technology predicts that the AI Singularity will occur by 2045.

Ray puts forward a number of convincing arguments in his book “The Singularity is Near” (Kurzweil, 2005) to support this prediction.

It would take a very long article to describe all of Kurzweils arguments, however the basic principle he puts forward is that development growth will be exponential rather than linear.  In fact he argues that not only is the growth of technology exponential but that “the rate of exponential growth (that is, the exponent) is itself growing exponentially” (Kurzweil, 2005) giving a double exponential growth.

Moore’s Law predicts similar exponential growth and irrespective of whether Moore’s Law is accurate or not, growth of computing power from 1965 to 2015 has been exponential.

Kurzweil says that it is this exponential growth in technology that will empower us to match the computational capacity of a human brain.

A counter argument has been put forward by Cirurel et al in their book “The Relativistic Brain” where they outline a number of reasons why the human brain cannot be simulated by a computer. Cirurel is a mathematician and Nicolelis is a neurophysiologist and they have spent a long time studying the brain and how it works.  By combining mathematical, computational, neurobiological and evolutionary arguments, they put forward a theory called the relativistic brain theory (RBT) which states that there are multiple networks of interconnected brain cells.  These generate complex brain functions such as sensing pain, sense of self and consciousness.  They dispute the idea known as computationalism, that the human brain resemble digital computers and can be reproduced or simulated by software running on a large computer.  There is a lot of money been spent on research throughout the world where large teams are working with state of the art computers to simulate animal or humans brains, yet despite this it appears that they are a long way from achieving anything that resembles a brain.  They argue that the reason for this is that there are non-computable and non-tractable problems that even the most sophisticated computers cannot deal with.

The RBT theory is that the brain is comprised of a hybrid digital-analogue computational engine.  The digital component is defined by spikes produced by the neural networks distributed all over the brain while the analog component is represented by superimposition of time-varying, neuronal electromagnetic fields (NENF’s) generated in the multitude of distributed loops of white matter in the human brain.  The relativistic brain theory proposes that “the combination of time-varying NEMF’s produces a neuronal space-time continuum.  This space-time continuum defines what we call the mental space where all higher brain functions emerge” (Cicurel & Nicolelis, 2015)

The variable elements of this are used to explain why we might respond differently to the same touch at different times.

They do agree that there are a number of motor functions in the brain that can be interpreted and simulated and there is ample evidence to prove this.  A good example is someone controlling a prosthetic limb using signals from the brain.

A more complex function is to trick the brain into believing it’s in pain, this can be done with visual stimulus (rubber hand test), but so far no one has been able to do it by sending a signal to the brain.

There is also an evolutionary argument which is that the brain developed as a result of a large number of evolutionary steps, through changes in everything from molecules to the entire organism.  Evolution does not follow any pre-set plan so then how can we expect to reverse engineer something that was never engineered in the first place.  (Cicurel & Nicolelis, 2015)

Kurzweil states that just as it was “the fate of bacteria to evolve into a technology-creating species”, “it’s our destiny now to evolve into the vast intelligence of the Singularity”.  (Kurzweil, 2005)

I propose that these arguments show that due to the complexity and as yet unknown functions of the human brain that the Singularity may a lot further away than some observers would like to admit!

References

Cicurel, R. & Nicolelis, M., 2015. The Relativistic Brain. Montreux: Kios Press.

Kurzweil, R., 2005. The Singularity is Near. s.l.:Viking Penguin.

Vinge, V., 1993. The Coming Technological Singularity: How to Survive in the Post-Human Era. Westlake, Ohio, NASA Conference Publication.

Advertisements

3 thoughts on “How near is the Singularity?

  1. Good Post!.
    I am glad that the human mind has barriers to replication. That is what protects us in a sense. A lot of software though seems to be able to bypass context and still produce accurate results, through learning and sheer processing power.

    Like

  2. Personally i think Kurzweil is wildly optimistic in his predictions. If we ever get to a stage where there is controllable Friendly AI then we can start to dream. I think my children may see it in their later years, I doubt if I’ll see it.

    Like

  3. I would agree that Kurzweil is very optimistic, Paul Allen has a strong opposing view. He believes that as we work on unravelling the complexity of the brain we will understand that its more difficult then first thought, he calls this the “the complexity brake.” He says that just because we can start to understand the brain does not mean we are even close to finishing the job!

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s