A Digital Brain by 2029

Last Updated on June 2, 2020 by

Scientists have pointed out that as the limitations of the integrated circuit limit the ability of an artificially intelligent machine, the projected progress outlined by Moores Law will ultimately finally reach the edge of its limitations.  But as new developments in quantum computers come forth each passing year, the AI of tomorrow may use quantum computers already in development to function.  And as such, what can we make of Moore’s Law which suggests a human digital brain could reach our desktops by the year 2029?

Perhaps the most interesting aspect of artificial intelligence is that Moore’s Law has not, as so many have predicted, slowed down.  Rather, in 2008 the progress of transistors in computers has -in fact- sped up.  In 1990 the release of the 486 computer carried with it approximately one million transistors and was capable of playing tic tac toe extremely well, but only capable of playing chess at a beginner level.  It has been estimated that the human brain’s cerebral cortex carries within it twenty billion neurons, and the entire brain houses somewhere around fifty billion.  If this analysis is to be believed, then the human brain is essentially equivalent in processing power to twenty-five Quad-Core Itanium Tukwila computer processors (developed in 2008).  Earlier this year a new type of transistor was developed that operated similarly to a human brain’s.  If we were to take the processing power of the most advanced processors and put them together with these new neuron-like transistors would we ultimately create a facsimile of the human brain?  Unfortunately, not yet.  Such a device would only have an equivalent processing power to a human brain and would not necessarily be able to process information in a comparable way.

But then we take into account the potential learning capabilities of social machines such as Kismet, which has a certain degree of interactive intelligence and even expressive features, and the powers of perception and reaction such as the TOPIO robot, which can play ping-pong, and somehow find a way for these machines to become creative using mathematical algorithms and suddenly the possibility of true Artificial Intelligence no longer becomes an issue left to the imagination of science fiction.  Instead, such a system would be able to interact with its environment in a way that would be difficult to distinguish from human interaction.  Additionally, such a system would almost certainly be able to then begin designing a new system of thinking that made programming as well as some forms of human problem solving a thing of the past.  It would then also at this point begin designing a more efficient means of improving itself.  At this point such a machine is at the threshold of reaching a singularity.  The future AI systems such a machine would design would be capable of designing better systems, and so on.

Current projections suggest we may have a realistic human brain in computer format by the year 2029, and this brain could be manufactured to help humanity at this level by not only conveying information, but developing and processing it.  And by that point will it be ethical to reprogram them or turn them off?

It’s easy when dealing with this issue to let fear be the primary guiding judge of whether or not to explore this new horizon of technology.  But if humans have developed the capability of creating life, are we not by this point capable of transcending such basic emotions as fear of the unknown?  And what could we learn about ourselves if we created an entirely artificial yet wholly conscious entity?