AI graduating from point solutions to enterprise-level systems
- As artificial intelligence becomes more ubiquitous, pioneers are having to shift strategy from enacting point solutions to enterprise-level systematic solutions, according to a survey of more than 3,000 business executives, managers and analysts conducted by The Boston Consulting Group and MIT Sloan Management Review.
- Microsoft, for example, is working on a "complete AI platform" to support "systems of intelligence," Joseph Sirosh, CTO of AI in the company's WW Commercial Business group, told surveyors. Such a system will include an "iterative learning loop" to support model creation, deployment and maintenance to continually improve business operations.
- But the gap between AI pioneers, the companies that have implemented it more extensively and are innovating at the edge, and companies just getting their toes wet in the technology is widening, according to the survey. While pioneers struggle with talent access and juggling other investment priorities as barriers to AI adoption, organizations with less AI experience and investment are more stunted by other technology capabilities, a lack of leadership support and unclear business cases for AI.
While point solutions are a natural first step for companies to dabble in a new technology, making the transition from older systems to new technologies can strain IT infrastructures without overarching capabilities to tie distinct parts of business together.
At Anheuser-Busch InBev, despite large troves of data that could serve as fodder for AI applications, too much time was going toward ingesting data and making it ready for use, Tassilo Festetics, VP of global solutions, told the surveyors. The company had to think of how they would structure data processes and platforms if they were starting from the ground up again, which allowed them to make data more available for AI applications.
To help distinct parts of business operate smoothly together, a common data platform and management system can help minimize data from going underused or ignored. And whoever puts the right data to work the fastest will more than likely emerge as the winner with the right combination of AI, data and the cloud to drive innovation.
Whether the company developing AI tools for clients or the company making internal data available for in-house or proprietary AI systems, there's a role to play for almost every organization in the race for AI capabilities. Most companies are pioneering, investigating or experimenting, but the latter two groups need to work to close the gap with pioneers.
Follow Alex Hickey on Twitter