When AI is bad for businessby
Kevin Philips explains why future-proofing your finance function with automation technology is a myth.
The industrial revolution brought about massive changes in how we work that rippled across society. Machinery replaced human labour in many aspects of the economy, from agriculture to manufacturing and, in doing so, changed the way these activities were carried out.
Our current artificial intelligence (AI) driven fourth industrial revolution seeks to do the same thing: with computers at the forefront.
The conversation around automation and artificial intelligence (AI) in the accounting world has been brewing for a few years (with a touch of panic about robots replacing us) and has been forced centre stage by the pandemic.
We need to support our organisations and clients during rapid and unpredictable change, and the case for automating our processes to speed up and improve our financial operations is stronger than ever before.
But the potential pitfall with AI and automation is that we run the risk of being too quick to automate existing activities to achieve quicker results without questioning the need for the activity in the first place.
Automation vs AI
While AI gets the headlines, automation is the foundation of any digitalisation strategy. The automation of mundane, repetitive tasks helps structure data in a way that enables an AI strategy. For AI to do its thing – learn and think in a way that resembles human intelligence – it needs to be trained on good, clean data in real-time. That’s where automation comes in.
Automation also offers benefits in its own right: speeding up manual processes, ensuring accuracy and flagging anomalies. Automation can ensure your data is always up-to-date and always where it should be – ideally in the cloud.
And using best-of-breed technology with open application programming interfaces (APIs) – basically computer-speak for making programs talk to each other – means your accounting software can seamlessly integrate and connect with software in other parts of the organisation, bringing further efficiency and accuracy.
But automation and technology can only get you so far if you use them to digitalise out-of-date processes that are no longer fit for purpose. Don’t use automation to make bad processes run faster because this only results in the wrong decisions being reached faster.
Consider whether your processes are laden with steps that are only there out of habit or because the process is being carried out by a human. As an example, if a process takes ten steps when completed by a human, consider whether a machine may be able to do it in two.
The future-proof myth
A year or more ago, I’d suggest that automating the right things in the right way future-proofs your business for the digital age. Today, that notion is laughable. The last year’s series of once-in-a-lifetime, outlier incidents occurring on an almost weekly basis made it impossible to carry out scenario planning.
Everything that has gone before, which would typically have informed future assumptions and decisions, while being historically true, is largely useless as being indicative of future trends.
Even AI, which is far better at spotting patterns than humans are, is going to have a very skewed perception of what is a trend. Data gathered in 2020 and 2021 is going to be far from typical. And, when we finally get to some level of stability in 2022, this ‘normal’ is unlikely to be picking up the thread from 2019’s ‘normal’. That’s four years of atypical data making it very difficult, if not impossible, to spot patterns, and predict trends and likely outcomes.
So future-proofing your accounting function with technology is a myth. But it is possible to future-enable it by using technology and rethinking the processes you automate. This allows you to build in flex to maximise opportunities and minimise the impact of challenges as they happen. And by doing so, it can put the accountant, augmented by technology, back in the role of trusted advisor to the wider business and clients.
We can add value by analysing change and predicting impacts across the business, now and into the future, which is exactly what we were trained for.
You might also be interested in
Kevin is the founder and CEO of idu Software. He has degrees in Commerce and Accounting, and started idu with partners James Smith and Wayne Claassen in 1998. Kevin is fast becoming a thought leader in his field, and makes regular comment in the media about current affairs affecting business...