Generative AI (GenAI) is transforming investment sector priorities and the future of human capital—as AI advances, so do jobs, skill sets, and workforce dynamics.
In our recent webinar, Charting the Course: A CFO's Guide to Modern Fund Management in 2024, Derek Shanahan, Strategic Partnerships Director at Juniper Square, spoke with Joseph Harrington and Dana Van Wie, both partners at PwC US, plus Adam Walker, Industry Principal, Fund Solutions, Juniper Square about the rapid development of artificial intelligence and best practices around deployment.
Despite the rapid rise and continued buzz around GenAI for business, the panelists didn’t predict huge shifts on the immediate horizon. Instead, they suggested that CFOs plan and prepare today’s businesses for eventual AI adoption. In the end, successfully leveraging AI will still rely on some almost old-fashioned business values—quality people and process management, smart data management and organization, and the right mindset.
People first still matters—more than ever
All the panelists agreed that adopting technology for technology’s sake tends to be a losing strategy. Instead, executives need to ensure that AI tools augment and support their people and processes in ways that are beneficial to the organization. Leading with business-first problems, however incremental or modest, is always the smarter course.
“I challenge my clients to think about where they are really struggling. What processes are really clunky and cumbersome? What roles do you have a hard time filling? Where do you feel like you're spending an exorbitant amount of time and energy wasted resources?” said Van Wie.
“It is really important to take a step back and be thoughtful about what makes sense in our environment, what makes sense for how we structure our current business processes, and where we are going with them,” she added.
Harrington agreed, noting that although AI development is moving fast, it's not reinventing the wheel overnight and the basics of the business, data, and processes may not change. Instead, the key lies in integrating GenAI into current processes to deliver incremental changes, which create small but critical efficiencies that can give the business a competitive edge.
The panelists also agreed that bringing your business needs to third-party consultants or vendors is the fastest, most efficient track for identifying the kinds of business problems GenAI is best suited to solve. These partners can also help you secure your networks so that controlled experimentation with AI doesn’t accidentally expose sensitive data externally.
Data hygiene becomes increasingly important
Clean, well-structured data is foundational, says Walker. “This all comes back to data hygiene and everything around it. If your data is not clean, most of these tools will typically be worthless. If you've got one initiative for this year…I’d focus that money and those resources on cleaning that data and getting it into a good spot.”
Van Wie recently met with a client struggling to assess why they were wasting so much time processing data and how they could better address the business challenges they faced. She encouraged her client to engage in the process of understanding where all their data lives and how it’s processed via their current systems—and then identify what tools would support streamlining and making more sense of that environment.
Hand in hand with business needs comes strategy and data governance, which Van Wie likens to seat belts: “Well-designed governance can really be analogized to seat belts and guardrails enabling us to drive faster but safer and not slower. If you try to start down the path [of adopting GenAI] without considering your data strategy and your governance policies for that data, you're only going to find yourself doubling back at the end of the day.”
In other words, teams can’t understand which tools make the most sense and will enhance their data hygiene if they don’t have an accurate picture of their current status and future needs.
With AI, the best will just get better
Harrington, a board member of the McIntire School of Commerce at the University of Virginia, shared an experiment from some of the university’s classes. For half a semester, the students leveraged no GenAI. For the second half, they used GenAI. The hypothesis was that GenAI would especially help struggling students.
That hypothesis proved wrong.
The better-performing students just got better. They were able to leverage AI to support their performance in ways that accelerated their growth, creating an even deeper chasm between the top and bottom students.
“The lesser-performing students couldn't review the output, couldn't impart any of their sort of human judgment or characteristics or analysis on top of it, and were just spewing out the answers from these models,” Harrington said. “The better-performing students would use [AI outputs] as a good starting point” to move past a mental block or enrich a perspective.
The implication here is pretty clear: When it comes to the private markets, the best minds in the business will likely find ways to use AI to augment their work, allowing them to move ahead even faster and more efficiently than before.
Prepare to iterate faster
Emerging tech moves fast, often faster than many tied to a CFO’s organization may be used to or feel comfortable with. The right mindset will emphasize flexibility, controlled experimentation, and data-driven outcomes to iterate incrementally toward the right solution.
“I tell my team in leadership—be ready to throw everything away in six months,” Harrington said. “The amount of change that's going to continue to proliferate is going to be a shock to many folks. Being prepared for change is the number one punchline,” he added.
Change is coming at us all fast, but there’s good news: Everyone is still learning how GenAI works and where its usefulness lies. Both the challenge and the opportunity of this current phase boils down to simply getting started. As Harrington noted: “The risk of inaction is sitting on the sidelines and really not doing anything.” He recommended interacting with Microsoft Copilots and other AI tools embedded into familiar software just to familiarize yourself with the tools, learn how to write more effective prompts, and control for hallucinations. Failing to do this is “missing the opportunity of practicing while everyone is learning.”