- Major Studies
- Market Assessments
Despite the buzz, artificial intelligence (AI) appears to be not finding as much traction in businesses as we would expect. On the surface, it seems most companies do not understand the use cases that would lead them to invest heavily in AI. But the story is more complicated than it first appears.
AI, as we define it, is not a single technology but several closely-grouped technologies that allow a system to absorb, analyze, and learn from data in order to make a recommendation or take an action. These technologies include rules-based reasoning, machine learning, speech recognition, natural language processing, facial/object recognition, and other capabilities. They are seldom deployed separately and are usually embedded as features or capabilities within business applications.
Some of these technologies are still searching for use cases. Outside of digital assistance and search, for example, natural language processing is not yet finding widespread business applications. But that may soon change, not just for natural language processing, but other AI capabilities as well.
As shown in Figure 2 from our full report, AI Adoption Trends and Customer Experience, only about 13% of companies that completed our Technology Trends 2020 survey reported that they have adopted any type of AI solution. However, investment is rising. This year, 27% of our respondents indicate that they are investing in new AI capabilities, an increase from 2019 when the investment rate was 24%.
Machine learning is another example of a technology that needs more use cases. In many companies, ML isn’t fully trusted in comparison to human experience and intuition. Another AI-associated technology, facial recognition, has also been slow to take off and find business applications. And many of those it has found are controversial for reasons of privacy.
“Some forms of AI are so ubiquitous we don’t even think of them as AI anymore, like the rules that move elevators in skyscrapers,” said Tom Dunlap, director of research for Computer Economics, a service of Avasant Research, based in Los Angeles. “But when it comes to the more sophisticated types of AI, use-cases are still rare. We expect that when more early adopters show success, adoption will expand quickly.”
Our full report provides an overview of AI adoption and investment trends, providing data on how many organizations have solutions in place, how many are in the process of implementing it, and how many are expanding implementations. We also look at the return on investment experience, total cost of ownership experience, and considered or planned uses for new AI investment. We conclude with important principles to apply in planning and implementing AI systems.