Countdown to the Singularity
January 9, 2015
In 2013 researchers from the RIKEN HPCI Program for Computational Life Sciences, the Okinawa Institute of Technology Graduate University (OIST) in Japan and Forschungszentrum Jülich in Germany carried out the largest general neuronal network simulation to date, exploiting the full computational power of the Japanese supercomputer, K Computer. ref
The simulation represents roughly 1% of neural capacity of the human brain and took 40 minutes to simulate 1 second of brain activity. That may seem rather paltry, but as computers continue their exponential growth under Moore’s law it easy to see that it won’t be long before they begin to approach human level capacities.
| 2014 | 2016 | 2018 | 2020 | 2022 | 2024 | 2026 | 2028 | 2030 | 2032 | |
|---|---|---|---|---|---|---|---|---|---|---|
| capacity | .01 | .02 | .04 | .08 | .16 | .32 | .64 | 1x | 2x | 5x |
| time | 40m | 20m | 10m | 5m | 150s | 75s | 38s | 19s | 10s | 5s |
As we can see from the chart, by 2032 a super computer will achieve parity with a human being. It isn’t hard to see the Singularity is but a few years away from then. At this rate, a computer will be a able to simulate the brain activity of the entire human race in real time by 2062.
Of course this assume the continuation of Moore’s law over the next 50 years and that software design is able to put the hardware to sufficient use. While Moore’s law is already showing some cracks, it is also clear that new technologies will ultimately compensate for the end of silicon miniaturization – the technology that has driven that law since the the invention of the integrated circuit. Just around the corner are new technologies including 3D chip stacking, racetrack memory and optical interconnects. These and a few other technologies appear to be enough to sustain growth until at least the 2030s and one suspects there will be new discoveries by then to move us even further, such as advances in spintronics and quantum computing. There is no reason to believe advancements in computation will come to a halt over the next century. At worst they may slow down, stretching advancements over a few additional decades, but no more.
Computers will surpass their makes in capability within the lifetimes of people alive today. We can only hope they will also surpass us in moral character as well.