Feature
“If non-A.I. software produces an inaccurate result, one debugs the program. When an AI produces an inaccurate result — say, for a classification problem — one tries to improve the output by increasing the training, feeding it more data, or changing some configuration patterns.”
‘Agentic A.I.’ will help organizations across the world achieve significant productivity gains, increase operating efficiency of existing resources, enhance creation of next gen employee/customer services and enable new business model development, said Harbert Executive in Residence and accomplished tech industry executive, David Parsons. Executives have been slow to bring the tools into the strategic discussion. Analytics have been around for a while and have gained traction, but only 7 percent of respondents to a recent McKinsey survey said they used AI in strategy formation. That’s compared to marketing, supply chain, and service, where the use is higher — 25-30 percent. Parsons again. “A.I. agents do more than simply assist. These A.I. agents will act intelligently to take action freeing people up to focus on the work that matters most, while fleets of ‘functional domain-use case’ specific A.I. Agents orchestrate end-to-end processes in every corner of an organization’s business, enabling true business transformation.” Every business probably has some opportunity to use A.I. more than it does today. The first thing to look at is the availability of data. Do you have performance data that can be organized in a systematic way? David Paradice, Harbert Eminent Scholar in Business Analytics and Information Systems, speculates that larger firms will initially have a significant advantage over smaller firms when using A.I. to formulate strategy. They have greater access to data resources to process more complex models, and gain granular insights that humans could not. Companies whose strategies rely on a few big decisions with limited data would get less from A.I. Likewise, those facing a lot of volatility and vulnerability to external events would benefit less than companies with controlled and systematic portfolios, although they could deploy A.I. to better predict those external events and identify what can and cannot be controlled. But even with vast amounts of data available,
- David Paradice Harbert Eminent Scholar Supply Chain Management Interim Chair Department of Business Analytics and Information Systems
determining which data is needed to train A.I. is a challenge in itself. An enormous range of information, encompassing industries, markets, economics, politics, transportation, even weather all come into play. Here, it might be smart to let A.I. serve as a subset of the strategy, with humans filling in the rest of the process. Parsons maintains that “the next gen technology platforms to build a Workflow Data Fabric to eliminate the existence of siloed legacy data are rapidly coming to market. [These platforms will remove] the issue of siloed data. The ability to connect all enterprise data from any source to ‘best in class’ workflow creation engines will be game changing.”
40 Harbert Magazine, Spring 2025
Powered by FlippingBook