A panel of decision experts, including Sparkling Logic CTO, Carlos Serrano-Morales, gathered at DecisionCAMP 2025 to discuss the impact of AI on business decision-making and work. You can find the recording of the whole discussion here. Here are highlights of what they had to say.
AI Cannot Replace the Basics
Most “solutions” fail because they lack a clear problem to solve. Solutions that incorporate AI are no different. It’s no wonder that a recent MIT NANDA study found that 95% of organizations that have invested in enterprise generative AI (GenAI) are getting zero return. The study found that common failure points include “brittle workflows, lack of contextual learning, and misalignment with day-to-day operations.” In other words, the approach rather than the models are to blame. Denis Gagné, CEO of Trisotech, encouraged organizations to go back to the traditional approach of solutioning: first define the problem and then assess the right technology. Carlos also emphasized the importance of measurement. If you don’t know what to measure and what your benchmarks are, then you don’t have clearly defined goals. Lack of metrics signals that there is something amiss in your solution.
AI Can’t Replace Soft Skills
While some roles and functions have already been replaced by AI, much work still requires the human touch. For example, designing a new decision is more than just writing business rules. Often this involves consensus building, commented Alan Fish, Senior Director of Business Enablement at FICO. Skilled business analysts know how to gather different leaders with different opinions on how the business should run and get them to come to an agreement on a decision model. Therefore, Alan believes that GenAI won’t be replacing Business Analysts anytime soon. These soft skills will be even more important as AI becomes more embedded in work. Carlos believes that skillsets will evolve from domain expertise to social expertise. Knowing how to ask the right questions will be key to successful human-human and human-AI collaboration.
AI is Expensive
“How much VC money can I waste today by asking ChatGPT to do addition for me?” joked James Taylor, Executive Partner and Founder of IBM Consultancy firm Blue Polaris. Today, most users do not realize how expensive GenAI is. Households near data centers are beginning to feel the cost of AI as their electricity bills have increased as much as 267% in the last 5 years according to a recent Bloomberg article. And data processing is only one part of the equation. According to Carlos, it is unlikely that organizations are going to rely on GenAI for straight-through processing (STP). To do so would require hiring a multitude of data scientists with PhDs to get the level of predictability, reliability, and explainability necessary. Even if cost wasn’t an issue, GenAI is often far too slow for high-volume transactions such as credit approvals, commented Gary Hallmark, Business Automation Architect at Oracle. As for non-STP decisions, Gary believes that “fully automated” is not the desired outcome for most organizations. Otherwise, organizations wouldn’t require sign-off from multiple VPs and stakeholders.
By going back to the basics, facilitated by business analysts with strong soft skills, organizations can determine whether or not AI costs are truly worth it.
Looking to automate decisions in a manner that keeps humans in the loop and in control? Contact us today for a customized demo of SMARTS™!
More in our DecisionCAMP ’25 Blog Series: