Pages

August 29, 2011

Perfect and optimized hindsight as a model for Post Singularity or any Enhanced Intelligence Model

I have previously provided my opinion on the three major views on what a possible Technological Singularity will be like.

1. Event Horizon - The claim is that future becomes weirder than we can imagine now and changes faster than we can handle. I believe that we will still be able to understand the changes but will be less clear on the details.

2. Intelligence Explosion - Core claim: Intelligence has always been the source of technology. If technology can significantly improve on human intelligence - create minds smarter than the smartest existing humans - then this closes the loop and creates a positive feedback cycle. What would humans with brain-computer interfaces do with their augmented intelligence? One good bet is that they’d design the next generation of brain-computer interfaces. Intelligence enhancement is a classic tipping point; the smarter you get, the more intelligence you can apply to making yourself even smarter.

3. Accelerating Change - Core claim: Our intuitions about change are linear; we expect roughly as much change as has occurred in the past over our own lifetimes. But technological change feeds on itself, and therefore accelerates. Change today is faster than it was 500 years ago, which in turn is faster than it was 5000 years ago. Our recent past is not a reliable guide to how much change we should expect in the future.

Intelligence explosion and Accelerating Change Approximation with Perfect Hindsight

Back in 2008, I wrote - Singularity perspectives using hindsight and optimal algorithms: AGI raised by wolves

We can use a thought experiment of placing a hypothetical superior intelligence back 20, 30 or 40 years or 2000 years, we could use our hindsight knowledge of superior algorithms/processes/methods (developed between then and now) and new technologies to approximate possible improvements a AGI could use. The level of advantage could be used to approximate advantages of a current or future AGI. It would give some indication of how advances could be accelerated to move from older technology to newer technology.

An optimal leapfrogging path of technology adoption could be determined. What would be the fastest path to go from some past state to what we know is possible now ? How fast could it have been done ? What would have to be built ? A fictional analogy is Spock in the classic Star Trek Episode City on the Edge of Forever. How fast can you progress ?

Spock - "I am endeavoring to construct a mnemonic memory circuit using stone knives and bear skins."

We will also have to update the timeframes and the capabilities based on the expected science and technology at the time that any Singularity occurs. For example if it were to happen in 30 years and in 5-15 years there is a massive adoption in robotics for personal assistance and manufacturing then that will be easier to accelerate.



Economic Growth

Looking at historical economic growth there were increases in economic growth rates because of the adoption of agriculture, early industrialization and Ford's assembly line methods. There are further optimizations that are possible to accelerate economic growth.

China has average at 9-11% economic growth for over 35 years by catching up and adopting proven technology and methods. Therefore a AGI boosted society should have accelerated innovation in new technology and methods to place that society into the equivalent of permanent catch up mode.

Useful things for any superintelligence to do if we have not done it yet

There is a list of technologies and capabilities that humanity has been trying to achieve for a while. It seems like any Superintelligence would want to have those capabilities if they were not already achieved.

Molecular Nanotechnology
Nuclear Fusion mastery
Advanced space capabilities

I believe that we will have those capabilities before true superintelligence, but it is possible they could come after.

Need to breakdown the curve projection

Many of the forecasts of the Technological Singularity are based on projecting out certain metrics (like computing power) on a graph and adding in some analysis of brain science progress.

Kurzweil does talk about S-curves for each particular technology making up an overall exponential progress chart.

Kurzweil estimate of brain capacity is 100 billion neurons times an average 1,000 connections per neuron (with the calculations taking place primarily in the connections) times 200 calculations per second. Although these estimates are conservatively high, one can find higher and lower estimates. However, even much higher (or lower) estimates by orders of magnitude only shift the prediction by a relatively small number of years.

Some prominent dates from this analysis include the following:

* We achieve one Human Brain capability (2 * 10^16 cps, 20 petaflops) for $1,000 around the year 2023.
* We achieve one Human Brain capability (2 * 10^16 cps) for one cent around the year 2037.
* We achieve one Human Race capability (2 * 10^26 cps) for $1,000 around the year 2049.
* We achieve one Human Race capability (2 * 10^26 cps) for one cent around the year 2059.


20 petaflops for $200 million should happen in 2012-2013

However, the usable processing power for brain emulation may not be at that level. The efficiency of brain emulation may not get to 1 to 1. A lot more processing will be needed for the computer system to


If you liked this article, please give it a quick review on ycombinator or StumbleUpon. Thanks

blog comments powered by Disqus