Generative AI tools like ChatGPT are still in the early stages of development, and it is not clear how these tools will change the finance function.
However the technology is implemented and operationalized, at the moment, there are not enough skilled people with the combination of finance and data skills to meet the change, Ashok Manthena, founder of ChatFin and a speaker and author on AI, told an audience at AICPA & CIMA’s CFO Conference on Wednesday May 10.
The AI future likely means the creation of a hybrid position, a combination of financial analyst and data analyst, within the finance function, according to Manthena.
“This is the role that is going to come in: finance data scientists,” he said.
Start studying. But the current data skill sets of finance professionals aren’t there yet, Manthena said.
“What I’ve seen in the last few years, because I tried to hire a lot of data scientists with finance knowledge. You can't really find any data scientists with finance knowledge,” he said. “The solution is finance people getting trained on data science. This has to be done.”
Finance professionals are best positioned to direct the future of AI within the finance function because they have the contextual knowledge and experience to cope with the complexity of financial data, regulations, accounting standards and methods, and organizational structure. The current state of AI technology isn’t capable of that yet, Manthena told the audience.
“We [finance professionals] all have this finance and domain expertise. Domain expertise is very important, which I think any large language model won’t have at this point of time” he said. “All the accounting rules, all the finance statement rules—the model doesn’t really understand a lot of these rules. But at some point…we need to make the model understand these rules as well.”
News built for finance pros
CFO Brew helps finance pros navigate their roles with insights into risk management, compliance, and strategy through our newsletter, virtual events, and digital guides.
Dream small. Finance professionals will need to train for, and adopt, the skill sets of data scientists to be most effective in an AI future, Manthena said. Finance professionals are interested in becoming more data science fluent, but it will take time to get there. For finance professionals looking to become more data savvy, he advised that they start small, with basic data skills first.
“I’ve seen a lot of interest coming from finance people. It will be a little tough, because you need to have a little bit of coding knowledge, like Python,” he advised suggested. “If you are curious, if you are enthusiastic, I really suggest starting with a basic course of Python, and then you can slowly go into data science.”
Manthena suggested that organizations should also focus on starting small, choosing AI projects with a medium level of complexity, like creating predictive analytics models for revenue or expense forecasting and keeping them separate from the larger organizational systems.
Then allow for testing and adjusting throughout the process before fully integrating into the system. Do it manually, he recommended. “Take your extract from your system, manually upload it into this model, get an output, use that output for a few months,” he said. “And then you can always integrate and automate this process.”—DA