The Gartner Hype Cycle for Artificial Intelligence (July 2019 – subscription required) has over 40 separate and distinct technologies that make up the AI market. Like most Gartner hype cycles, the one for artificial intelligence is heavy on the “Innovation Trigger” phase (phase 1), reflecting the enormous amount of innovation coming out of start-ups and the pace of that innovation; many technologies are in the “Peak of Inflated Expectations phase” (Phase 2 – the zenith of the cycle) and then quite a number in the grandly entitled “Trough of Disillusionment” – the third phase in the cycle. These include well-known AI technologies that have entered common parlances such as Machine Learning, Natural Language Processing, and Autonomous Vehicles. However, there are no technologies in the “Slope of Enlightenment” (phase 4) and just two in the “Plateau of Productivity” – the final phase (those two are speech recognition and GPU accelerators).
If that all made little sense to you, or you are not familiar with the Gartner hype cycle construct, fear not. It just means that from Gartner’s viewpoint, most AI is still pretty immature and a long way from reaching any form of productivity for organizations. This may well be an accurate reflection of the state of AI from a holistic perspective but is not necessarily a reflection of the work most of our enterprise clients are doing. Indeed, several recent studies have focused on the impact that AI is having on business and come to the conclusion that AI is indeed delivering meaningful results to those forward-thinking enough to invest.
One such recent report is from the esteemed McKinsey Analytics team and the title of the report tells its own story, “Global AI Survey: AI proves its worth, but few scale impact”. To distill an extensive piece of work down to a couple of sentences, it says that AI adoption is increasing (no surprise), most adopters are seeing positive results (that’s good) but high-performers, those who are essentially “all in” on AI, are seeing disproportionately higher returns in both their top and bottom lines (very encouraging).
How are the high performers (those that have adopted AI in more than 5 business activities) “all in”? It appears to come down to how well AI is supporting the overall business goals, the collaboration across functions for use of AI, and how much investment the company puts into training and re-training its workforce to use the technology. If the company ties its AI into its operating strategy and sees it as a means of upskilling its employees, then the gains are exponentially higher than if it sees AI as a standalone project. This is the “scale” that is referred to in the title of the report. Those that focus this technology toward the holy trinity of making money, saving money and managing risk are seeing the benefits. This is no surprise to us at Seal. Our clients are by and large looking for technology to solve specific business problems or take advantage of opportunities that they otherwise couldn’t take. The fact that the technology is based on algorithms, uses AI models and does clever things to extract text from documents is nice to know, but not as nice as delivering an ROI. Ultimately it comes down to economics, as it often does.
There is no denying that there is a considerable amount of hype around AI, just as there was several years ago with Cloud computing and before that the cadre of “as a Service” technologies. As with those previous paradigm shifts, the temptation to “AI-wash” current tech to claim an AI badge is too hard to resist for some, and as we have said often on the blog, not all AI is the same, the devil is in the detail. However, McKinsey’s report makes it clear, high performing companies, those who are “all in” on AI are seeing very tangible results.