You can’t always get what you want
But if you try sometime you find
You get what you need
– Mick Jagger, Rolling Stones (1969)
Every age comes to, what feels like, a crashing end. Whether it was the end of the swinging 60s as embodied by above lyrics written by Mick Jagger, or this pandemic that has gripped the world as we enter the 20th year of the second millennium.
The end is cathartic, but there are notes of uplifting change, just like the choir in this Rolling Stones’ track. We’ve learned a few things over the past decade that will allow us to carry on into the next age. We will get what we need, even when it wasn’t everything we wanted.
The same goes for accounting firms and the office of finance. The past 10 years of digital transformation had allowed most firms to be ready for COVID-19, while the remainder have had no choice but to catch up in the most painful fashion over the last 3 months. Last decade’s leadership in digital transformation has laid the groundwork for the next age, the Forth Industrial Revolution, which will be driven by Artificial Intelligence (AI).
CFOs and accounting firms who had completed the digital move early have all been exposed to the “footloose men” of AI in the last decade, experiencing the hype that AI would bring to the modernization of accounting and the office of finance. Most raw AI tools, such as Jupyter notebooks are rough frameworks, requiring firms to hire $250K/year data scientists to load and review data to only find out that they were missing the data needed to answer the financial question they wanted to ask in the first place.
The problem was that these “footloose men” sold AI and the associated tools as a single silver bullet, when really, it’s a mix of tools and skills. Gartner has been tracking the hype cycle around the tools and skills required to implement AI effectively, tracking what appears valuable, what is actually providing value.
Three of the key skills that Gartner just started tracking, AutoML, Augmented Intelligence and Explainable AI, are tools and skills added to the hype cycle that will be of value to accountants and the office of finance very early on in this decade. These skills are going to be critical in boosting the performance of the office of finance, or as an external accounting firm, providing the right advisory services to your clients.
Gartner describes AutoML, or Automated Machine Learning as the capability of automating the process of building, deploying, and managing machine learning models. It is closely tied into the AI Cloud services tools (also on the hype chart), whereby accountants and finance analysts can easily load certain types of financial data into a cloud-based service and have the system perform a 100% data review using the latest machine learning algorithms. The beauty of AutoML and AI Cloud services is that it’s a production grade environment, using proven and tested service models backed up by a team of support staff that can understand your financial operational environment.
Which brings up a skill not explicitly listed on the chart, but which is incredibly important, which is data literacy. In order to use AutoML, you have to know how to get the prescribed data out of your financial system in the right format for analysis. Often this is a job left to IT professionals, however IT professionals often lack the specialized accounting training to know if the data set extracted is complete or not. As a result, the actual work, known as extract, transform and load (or ETL) becomes a long back and forth process between accountants and IT professionals. Having that skill set in the office of finance or provided by your AI Cloud service provider is critical to seeing the benefits of AutoML at the speed they are intended.
Gartner describes augmented intelligence as “a human-centered partnership model of people and artificial intelligence (AI) working together to enhance cognitive performance, including learning, decision making, and new experiences.” For accountants and CFOs, it’s about finding a way of integrating AI into the way they work on a daily basis to take full advantage of it’s capabilities to boost their corporate performance.
This sounds far easier than it is. Training staff on AI tools and financial data extraction is one thing, but overcoming the fear that AI will replace staff once integrated into the procedures of the company is far more difficult. The silver bullet AI salesmen, and to some degree the Hollywood vision of Terminators coming after us, leads many to believe AI is days away from replacing professionals. The quick answer is that it won’t replace professionals, and likely never will.
AI is very good at analyzing large amounts of data very quickly, however only through the professional lens can you definitively determine if something flagged as “risky” by the AI is really just a routine transaction, or one that is well understood. It’s this human-machine partnership between AI and accountants that is going to be critical to drive the first gains in the office of the CFO. Additional AI tools, such as reinforced learning, could help to inject more accounting smarts into AI models, but in the end it will still require a trained professional to validate the result.
And part of that new set of professional skills that will further empower augmented intelligence is Explainable AI. Explainable AI is the reaction to hangover caused by the silver bullet AI salesmen who would tell you that if you load the data, the AI would magically give you a result. When you ask how it works, their answer would be to say that even they weren’t smart enough to know.
This black-box approach to AI was making it difficult for professionals to know what was contributing to the result being produced by the system. In the case of accounting professionals, a risk score calculated by an AI system against a transaction would be hard to evaluate as useful without an understanding of what analysis was taking place within the AI.
AI researchers define Explainable AI as an ensemble of methods that make black-box AI algorithms’ outputs sufficiently understandable. It will be key for accountants and CFOs to become literate on the different types of algorithms and what they do to better interpret results. For example, knowing that an AI uses a mix of algorithms that flag suspicious keywords and rarely used accounts can help an accountant understand the value of a risk result in trying to find possible fraud vs. looking for risky vendors who are delivering goods without a PO.
Cue that heavenly choir! Vigilant AI offers the latest AutoML Cloud AI based services, and our staff are fully versed in augmented intelligence and explainable AI. Whether you are in the office of the CFO looking for a way to quickly answer a pressing financial question, or if you are an accounting firm looking to add depth in advisory services, Vigilant AI has the key tools accountants need to allow you to thrive in the Fourth Industrial Revolution.
Vigilant AI resources can be contracted for one-time financial analyses or can be contracted to manage the data extraction and calculation of recurring analytics and dashboards for continuing use within your business. We will help your office answer your most pressing financial issue using the latest AI driven technology.
To arrange an online meeting, visit our Contact Us page.