Most futurist experts believe that Artificial Intelligence (AI) is going to play a key role in the development of technology. Some of those believe that AI will ultimately provide the quickest route to a ‘technological singularity’, past which no further technological innovation will be possible. This article will use the rich picture we have built, as well as some of the considerations we have made in previous articles, to assess the possibilities for the future of AI.
AI in virtual personalities
We’ve asserted previously that virtual personalities enable users to interact in a contented manner with multi-function devices. Critical to the development of virtual personalities is the evolution of AI. To accurately simulate anthropic entities, these virtual personalities require a degree of autonomy similar to human personalities. That is, if asked a question, the virtual personality must be at least capable of replying in a similar variety of ways.
This is a trickier problem than it might seem. Though most questions asked of people are guided in to restrictive ‘answering frames’ that limit the number of possible responses by individual’s emotional reference points, the nature of selecting the right ‘frame’ is complicated. So, AI must be developed not only as a logical engine running a program to discriminate human from non-human responses – the quickest route to effective AI seems to be in simulating the human brain itself.
Futurist Ray Kurzweil believes this will be achieved by 2020. That’s eight years’ time. At that point, what can we expect?
As we discussed in the previous article, companies will need to develop ‘avatar personalities’ that represent the characteristics of their corporate ethos. In doing this, then, they will need to really get to grips with the human psychology behind developing ‘human’ AIs. So, I anticipate a boom in research fields around neuroscience, cognitive psychology and, eventually, ‘machine psychology’ – the art of coding the human brain in to a logical device. Through careful study, corporate entities will be able to create machine personalities that are representative of their defining characteristics. Governmental entities may also pursue this line of thinking, but (as evidenced historically) on a slightly less progressive timeline.
Recall that consumers are, in general, driving towards a single unified ecosystem of content and procedures. This is mirrored in various global drives for integration and ‘end-to-end’ control. A company that produces its own printer now make its own paper. It’s possible to run your whole life using only products and services from Virgin.
What does this mean for the development of AIs? There are two clear paths here: a singular common ecosystem is established before the development of ‘human-level’ AI, or a singular common ecosystem is not established at all. In the second case, one could imagine cross-compatibility to be of little relevance: any cross-functionality will be achieved via interactions of high-level AI, consolidating the need for various communication frameworks in a common, natural, language. The first case is, to my mind, less likely (given the timescales involved). We have seen that there are multiple large entities pursuing the R&D dream – and all are succeeding to various extents. Market competition prevents them from collapsing together to form a single entity, and patent protection prevents their innovation from following identical lines. So, it is this author’s belief that the future is one of multifarious and multipersonalitied AIs communicating via a common, natural, language. At that point, I suspect the distinction between ‘real’ and ‘virtual’ personalities to be much, much less of an issue.