There is a wide range of views presented in the literature about how soon artificial intelligence techniques will have a significant impact on industrial, military and commercial businesses. At one extreme is the view that artificial intelligence will remain confined to the academic world with the emphasis on replicating various aspects of human intelligence but we are already seeing many influences and implications in industries at large of Artificial Intelligence agents in the wild. At the other extreme is the position that artificial intelligence is already revolutionizing the way business gets done. It is now at a point where it is impossible to deny that Artificial Intelligence is changing the way we work, the ways that we interact and the ways that we live. I will provide some examples and benchmarks in order to assist you, the reader, in evaluating the timing of the maturation of these technologies. Based on the examples in my previous posts, it seems as though artificial intelligence technology is actually moving quite rapidly into commercially viable applications. There is an interesting parallel to the genetic engineering industry. Impressive results in university laboratories triggered a rash of new ventures and claims for instantly building large business entities which would be exceptionally profitable. This was soon followed by disillusionment and increasing government attention to the social implications. Finally, however, genetic engineering is taking its place as an important new industry.
As a second example, much of the early and significant research in artificial intelligence was oriented to medical applications. As yet, however, there are few systems in widespread use. Although impressive performance has been demonstrated by a few large diagnostic systems built as research programs, few practicing physicians utilize AI-based systems. This may be due to front-end costs, mistrust of claimed performance, unwillingness to delegate responsibility, liability concerns, impersonal interactions, etc. Perhaps some of these concerns will be mitigated by the changing economics of medical practice or the mounting possibilities for malpractice in the space.
It has become clear over the past two decades that academic endeavors are deliberately driving commercial applications of their discoveries. Any discovery that could spawn a new product or service is ready to be spun into a corporation that seeks funding to further those prospects and AI is certainly no different. As natural language processing engines matured, dozens of different philosophical iterations of these parsers have been built and put into commercial use to drive the current chatbot revolution.
Establishing the commercial viability of an AI-based product can be difficult. It can be influenced by a number of factors including:
1. Appropriate marketing (including distribution and support)
2. Acceptance by the user community of a different way of doing things
3. Better performance or capabilities than present techniques based on a user-oriented product design
4. Validation of performance
5. Acceptable initial and ongoing costs
6. Effective user interface
7. Competitive or peer pressure
8. Status of the supplier
9. Ability to mimic human scale interactions with AI interactions