ChatGPT is just the start. AI will replace the equivalent of 300 million jobs. Most people aren’t ready for that. Goldman Sachs for one predicts that over 40% of all administrative tasks will be automated.
The finance function is far from immune from this disruption. If anything, we are on the cusp of a transformation in finance. We are lucky enough to have insights from a true thought leader in finance to analyse this change.
A Fireside Chat With Glenn Hopper
Glenn Hopper is a CFO and author of Deep Finance: Corporate Finance in the Information Age. The book is a must-read for CFOs, all of whom are facing challenges with emerging and extant technology. Our focus at Quolum is solving finance challenges around managing the legion of SaaS tools businesses use today, so we were very excited to have the chance to sit down with Glenn for a discussion about the hottest topic in SaaS and finance (and every other industry): AI.
Our chat with Glenn was so interesting we couldn’t help but share some of the highlights of it here; but if you find this post interesting, please take a listen to the full discussion with Glenn below.
Our chat kicked off with a doozie: While Glenn recalls milestones in technology like moving from a typewriter to a PC, the early days of the Web, and the rise of SaaS and machine learning, he contends that what we’re seeing in AI today marks the start of a much bigger transformation. The rest of our session focused on the particular implications of this transformation for finance.
The history of automation in finance
For years, finance professionals have used automation tools to make their lives easier—to be a master of Excel is to be a master of macros. AI won’t build on this traditional automation. It will supplant it. But AI isn’t the first technology to disrupt finance activities: After the first generation of desktop-based automation tools, the rise of SaaS tools moved automation to the cloud, where a firm’s data could be aggregated with vast amounts of external data. This development gave rise to the data-driven organization.
Finance in a data-driven organization quickly found far more decision-making power with organization-wide and even industry-wide insights. Later, we saw the rise of machine learning, which only became useful when data volumes reached a tipping point that allowed it to generate useful information. While these technologies have helped make finance more efficient, AI will drive automation further up the ladder of role complexity.
From automated tasks to automated roles
With few exceptions, pre-AI automation in finance made life easier by reducing routine and menial tasks (of which finance offers many). In the short run, AI will enhance this capability, but in the longer term AI will begin taking over more complex roles like tax and SEC filings. Eventually, it makes sense that AI will be able to fill even more complex functions like data analytics, financial forecasting (which could become real-time), and fraud detection.
Ultimately, no finance role is immune from optimization or takeover by AI tools, but it’s unclear whether advancing AI will take finance jobs in the direction of specialization or redundancy. For example, AI could have a role in accounts receivable tracking slow-to-pay customers. 80-90% of accounts receivable tasks may be delegated to AI in this scenario, but will it mean the loss of jobs or free up finance staff to spend more time managing the most critical billing situations?
SMBs vs. enterprises
One element we discussed in our chat was whether the availability of AI technologies would affect SMBs and enterprises equally. As with machine learning, there’s a tipping point for AI below which it becomes difficult to derive value from the technology. Enterprises almost always have access to more data for AI to ingest, but the case for adopting AI tools may be more difficult to assess for SMBs.
Though, there are tools in development to cater to SMBs like startups, which often have few finance staff. These tools will be able to walk non-finance SMB leaders through the core finance decisions they should be considering in plain language—the same way a human finance pro on the payroll would today.
Services that aggregate industry-wide data also make sense for mom-and-pop businesses that aren’t interested in or capable of being data-driven, but could benefit from aggregate insights. For example, a laundromat might not have the data to fuel an AI decision on where to best place a second business location, but a service that aggregates data from thousands of laundromats could offer enterprise-scale insights to even the smallest businesses.
AI and Finance: The Brass Tacks
AI is here, it’s advancing, and it’s poised to change everything. While it’s fun to speculate on exactly what that change will look like, we can’t be confidently predictive at the moment. To address emerging AI tools, we must have a framework for action. Glenn suggests firms ask themselves three key questions:
- What is this change?
Particularly as AI as a practical tool for business has appeared so quickly and is advancing so rapidly, we must be precise in our understanding of the particular tool, service, or technology we’re considering. - How can I exploit the change?
One way of looking at this is comparing a tool’s capabilities against an intended task’s position on the Harvard Business Review risk/demand matrix. In other words, how bad will it be if AI gets things wrong, and how much work would AI save us? - How can I convince others?
These are just the highlights of our chat with Glenn Hopper. In our full discussion, we take a deep dive with Glenn into more philosophical areas, like how AI will affect employment and the social and political implications of that (will we ever see a universal basic income?), as well as the fraught and complex topic of AI ethics (who should design AI guardrails?).
If you found this post interesting, please review our full discussion with Glenn Hopper here.