Follow The LIDA - A Software Bot With Conscious Processes
German philosopher Sir Christopher Riegel first proposed the existence of the unconscious back in the 18th century. As the concept developed, much of the focus went to the contrasting conscious mind, and the possibility that it was the thing that set humans apart from the rest of the animal kingdom. While debate continues on this matter, the assumption that computers could not possess consciousness has largely held firm.
But now, even this seemingly self-evident notion is being questioned. Researchers from the Cognitive Computing Research Group (CCRG) at the University of Memphis have have been developing an Artificial General Intelligence (AGI) software bot with the intention of having it develop conscious processes.
Known as the Learning Intelligent Distribution Agent (LIDA), the bot has recently demonstrated what the team considers to be consciousness, at least according to Global Workspace Theory (GWT). GWT posits that consciousness comes from the 'broadcast' of previously learned unconscious processes to the mind's 'global workspace.'
The team, led by Dr Stan Franklin, has now developed the bot to a point where it is just .08 of a second slower than the average person on a reaction time test, and to where it experienced the same illusion as people did in a perception test. For each of these tests, the bot first had to understand what it was 'seeing' before it could process that information in the 'conscious' manner required by the tests.
These results place LIDA at the front of the AGI field, which is one reason that CCRG are suggesting that the LIDA framework be adopted as the standard for cognitive architecture. As you may well imagine, this sort of technology is incredibly complex, so developing a standard will minimize redundancy in the development of this kind of software. Currently, there are at least 29 different frameworks out there, which means that a lot of brainpower and expertise is being wasted on achieving similar results.
Here is Franklin's colleague, David Friedlander, explaining the theory behind LIDA: