My last article of 2017 was a recap of AI World, and my first article of 2018 was about the developments in AI technology that have been achieved so far. Now I’d like to close out my conversation about artificial intelligence (AI) by recapping some of the notes from the presentations and talking about where the presenters believe things are headed.
One such presentation was “Making Robots Intelligent, Cognitive and Aware,” with iRobot CEO Colin Angle and Jibo Founder and Chief Scientist Cynthia Breazeal as presenters. They wanted to make clear that the robotics industry is still in the earliest of stages and still has a long way to go to achieve its primary goal. What is the first thing that comes to your mind when you consider the primary goal of robotics? Well, the primary goal of robotics is to extend people’s ability to live independently. Helping take care of older retirees and facilitating their ability to live in their own home without ever having to leave or being dependent on any humans for their care should lead to better quality of life in the later years.
Personal robotics has emerged as the current theme and as such brings a quest for a social robot. These robots will be a partner to interact with. Social robots will be made to be companionlike in nature. There will be different robots for different demographics, each of which will bring along a different paradigm and relationship with the device. These social robots of the future will have an ability to understand emotions and facial cues.
In the meantime, the robotic vacuum market is growing at a twenty percent compound annual growth rate (CAGR). Currently, this appears to be a bit of an obsession in the US, where there is nine to ten percent market saturation for the Roomba. (As a side note, did you know that somewhere around 80 to 90% of the people who get a Roomba actually name it something?) Of course, the roadmap for the robotics industry does not stop with vacuums.
Gartner VP and Fellow Tom Austin gave a presentation titled “AI Reality Therapy,” in which he shared that he is impressed with the actual science of AI and how it has been applied in the areas of economics, business innovation, and deep neural networks, but not so much in the actual applications. When speaking about the science, Mr. Austin made the point that AI is still a very hard thing to do and is not a repeatable science yet.
There will be multiple AI vendors in the future, and this technology should not be something that is dominated by just a couple of providers. So, when does the science fiction AI become a reality? According to Mr. Austin, we need around five more generations of AI-enriched applications. Maybe in the next fifteen to twenty-five years the science fiction will become a reality. Meanwhile, Mr. Austin recommends that companies need to be fast followers when it comes to utilizing AI. Companies should seek out tactical victories by looking for applications that meet the company objectives. Mr. Austin also believes there is far too much focus on the building of the AI, and that now the focus needs to move more toward the development of applications that can take advantage of AI.
Mr. Austin also shared that currently only somewhere around four percent of the enterprises surveyed by Gartner have some form of AI in production. Another four percent of the enterprises surveyed have claimed to have AI in the works or in development, while another fifty percent of the respondents are currently investing in experimentation within research and development. When it comes to current usage, about 60 to 70% of AI-enabled applications are focused on customer-facing applications.
In closing, artificial intelligence is still in its infancy, and we have quite a few more hurdles to jump before we get anywhere close to the primary goal of AI. I hope I have the opportunity to live long enough to see the science fiction AI become a modern-day reality, based on what these presenters have to say.