A perception gulf between science fiction and business fact exists in AI. This split is primarily the result of popular culture and the ever-busy technology hype machine. AI can be the stuff of science fiction, where machines’ intelligence outpaces that of their human creators, often resulting in world-as-we-know-it-ending apocalypse. The enterprise reality of AI is both less terrifying and more enriching for enterprises who adopt it: a labor-saving, outcome-improving technology that enables productivity gains.
Embedding AI into enterprise applications will enable productivity gains while tackling the adoption headache
The majority of currently realistic AI capabilities are task-specific, data-trained intelligence that acts as a copilot for the user, better informing them with timely, in-context information or suggested actions, and ultimately leading to automating (within defined limits) mundane tasks that a human might otherwise have wasted time on. This narrow, task-specific AI is where business value can be found for enterprises, and that value is primarily a function of enabling improved productivity.
Without getting too deep into macroeconomics, labor productivity is equal to a ratio between an output volume measure (the work completed) and a labor input measure (how much effort was required for the output). Therefore, with AI-powered capabilities focused on helping users complete their work more efficiently, we can reduce labor input (e.g., reducing manual effort to complete a task), and/or increase the output volume (e.g., through better outcomes enabled by data insights). In both cases, the gain is found in increased labor productivity – a key measure of performance for both enterprises and national economies, and one that has been determinedly sluggish in some sectors.
The key to unlocking these gains is not the technology itself: it is getting users to adopt it. Fortunately, a shortcut is emerging courtesy of the enterprise applications vendors, who are directly baking AI into core enterprise applications. Embedding AI capabilities into an enterprise application solves three big adoption headaches. First, it places AI in the users’ “go to” apps, no need to switch between screens or solutions while granting user type context to personalize AI interactions. Second, it puts AI use in the context of already established business processes, reducing the need for any immediate change management. Third, it sits AI capabilities directly on the data source they require to feed them, although effort to ensure quality and completeness of the data is still required. This approach is already picking up pace, and Ovum expects to see an ever-growing range of AI-powered enhancements to enterprise applications. One big cultural challenge remains, users will take time to trust their new AI co-pilots – something that will not happen overnight.
Straight Talk is a weekly briefing from the desk of the Chief Research Officer. To receive this newsletter by email, please contact us.