AI ‘perfect for the heavy lifting on routine tasks’
Many functions, such email queries or picking up anomalies in returns, lend themselves to the technology says Damien Greathead of Intuit Quickbooks.
Artificial intelligence can do the heavy lifting in many tasks and the more we use it, the better we will get at making the most of it, says the leader of the accountant and adviser group at Intuit Quickbooks, Damien Greathead.
“We will learn how to better define what we're looking for,” Mr Greathead said on the lastest Accountants Daily podcast. “Also, every time I use it, I become better and better at the instructions that I'm providing to it.”
Even with an imperfect result, the technology offered a leg-up in many tasks.
He said one recent example involved writing job descriptions to recruit three people to his team.
“I put it into ChatGPT to create that job description for me. So rather than me trawling through Google to find examples and building my own, I plugged it in and all of a sudden, 30 seconds later, that job description comes up.
“Ninety per cent of the heavy lifting has been done so I actually go straight into editing mode, rather than creation mode.”
Examples specific to accounting where AI could give staff a head start included routine filings such as BAS.
“There's an enormous amount of double-checking, triple-checking of activities, balancing … do transactions have a source document? Are there duplicate transactions?”
“Are there any anomalies, so for example we haven't paid this supplier in 12 months – why all of a sudden, are we sending a payment out to them?
“These transactions don't actually have a source document sitting alongside them, so we might want to query them. So there’s a bit of fraud prevention happening there as well.
“And this typically took a set of eyes to do month in, month out – if we had the time.
“Whereas if we allow the technology to do it, the technology flags it, we check it, review it and then move on.”
“There's a whole bunch of other examples that we're starting to see of using, say, ChatGPT capability in an accounting firm as well.”
He said those who misused AI – at the moment, at least – were being rapidly exposed.
“Look at that case in the United States where the lawyers used ChatGPT to build their case and create their closing arguments – they got found out pretty quickly.
“We're also seeing it in education as well. For all of the platforms that enable a student to use ChatGPT to create an assignment, there are also the platforms that analyse what is being submitted by students.
“Previously, you used to have to submit your assignment to make sure that it wasn't plagiarised, whereas now you'll be submitting them into these ChatGPT engines to make sure that they haven't been ChatGPT generated.
“So there are risks, obviously, of using it and I think you if you're trying to get to the point of doing as little as possible and leveraging as AI as much as possible, you’ll probably get caught out fairly early.”
He said many repeatable office tasks lent themselves to time-saving using AI.
“Think about the volume of email that comes into an accounting firm and how many of those emails are asking about things like, What's the status of this? When can I expect my return? How much do I owe the government?”
He said some accountants, such as the Carbon Group, were already working on ways to connect email with practice management software to generate draft responses.
“So rather than you the accountant or the firm administrator doing these types of activities, the machine, the inbox, receives the email, identifies it as a client, identifies it as a query, and then runs the query on the practice management database to see what the actual status is and what the last note was, creates the response for you.
“If 30 per cent of email is related to these types of queries, how much of the firm's time would be freed up?”
About the author
