In 2013, Oxford University released a research paper about Artificial Intelligence and automation. In this paper they stated that there is a 95% probability that all accounting jobs would be lost (and a 99% probability that all tax related accounting jobs would be lost). This research paper sent a shockwave through the accounting community, and people continue to reference their findings. We are going to take a look at what has happened since this paper was published and what this means for the future of accounting. I will also give you a different perspective and explain why Artificial Intelligence CANNOT automate accounting jobs.
I get asked a lot of questions on this YouTube channel. The question I get asked the most is, will accounting jobs be automated in the future, and should I pick a different career? There is a huge fear in the accounting community, that all accountants are going to lose their jobs to computers. This fear all comes from this one paper in 2013 from Oxford University. It is crazy to see the impact this paper has had on the accounting industry.
Here is the paper. “The Future of Employment: How Susceptible are Jobs to Computerisation?”
This was written by Carl Benedikt Frey, an Economist who studies automation, and Michael A. Osborne, an Artificial Intelligence Software Engineer. It is important to realize that neither of these authors are accountants, and they probably have no idea what an accountant actually does.
In this paper, they create a list of 702 different occupations ranked on probability of being automated by computers. Basically, if you ranked high on the list, you will likely lose your job within the next decade or two and be replaced with a computer. Accounting, auditing, and tax ranked very high on the list with accountants and auditors having a 95% probability of being automated, and tax accountants having a 99% probability of being automated. Overall, the paper claims that 45% of the US job market would lose their jobs over the next decade or two.
This paper was written in 2013 and we are now eight years later in 2021. 45% of US jobs have not disappeared. What happened? The paper obviously made some mistakes somewhere. Let us look a little closer at the analysis.
The paper uses a list of 702 occupations and evaluated each one on two main criteria:
- They asked the question, “Can the tasks of this job be sufficiently specified, conditional on the availability of big data, to be performed by state of the art computer-controlled equipment?” Basically, is the job a set of repeatable tasks you can automate?
- They did an additional analysis to determine if the tasks were the type of tasks that would be difficult for a computer to perform. They looked at three areas:
- Social intelligence – do you have to interact and persuade others?
- Creativity – does it involve unique problem solving?
- Perception and manipulation – does it require coordinated movement?
Based on these areas they made a judgement call. They looked at accounting and said that accounting jobs are full of repeatable tasks, that require no social intelligence or creativity. And then they assigned accounting jobs a score of “1” (which stands for the highest level of automation potential). With this high score, they then ran through probability calculations and came up with the result that 95% of accountants will lose their jobs and 99% of tax accountants will lose their jobs.
Unfortunately, this analysis is not correct. These researchers clearly do not know what an accountant actually does. For them to say that accounting is just a set of repeatable tasks that requires no social intelligence and no creativity shows they do not know what an accountant does.
In fact, if we look at what has happened since this paper was written, the number of accounting jobs in the US has actually grown. If you go on the website for the US Bureau of Labor Statistics, they predict that “employment of accountants and auditors is projected to grow 4 percent from 2019 to 2029” (which is expected to follow the growth of the economy).
I am focusing on this one research paper, but there have been a number of papers that have been written that all say the same thing. They identify accounting as a prime area for automation to replace humans with computers. Almost all of these papers suffer from the same problem. They are almost never written by accountants, and they misidentify what an accountant actually does.
What do these papers miss? I am going to give you three main reasons why Artificial Intelligence cannot automate accounting jobs.
- The point of accounting is to watch the money. You need to have a set of human eyeballs. Unfortunately, the business world is full of shady people. If employees have the opportunity, they will often walk off with your money. If you have an accounting department, you have a group of trustworthy people throughout your organization that is watching the money and making sure that none of it walks off. To use an analogy, if you were running a bank, even if you had a high-tech computerized security system, you would still have an armed guard on-site in case the computer system goes down. The same thing happens in a business. If you were to get rid of your accountants and replace them with computers, it is a lot easier for your employees to steal money from a computer than from a good accountant. Let me give you an example. Imagine you had an employee take a sales trip to California. Now imagine that sales person, while they are in California, decides to take a trip to Disneyland on the company credit card. If you have an accountant watching the expense reports, they will instantly catch that your employee is stealing your money to pay for their personal trip to Disneyland. A computer is not going to know the difference between the Disneyland tickets and a legitimate business trip expense. That is just a simple example. The point is that accountants are not just sitting at their desks performing mindless processes. They are your first line of defense against your employees stealing all your money, and I think you are a fool if you hand over that responsibility to a computer.
- Software engineers have an agenda. You have to ask why accountants are singled out in these research papers. Out of all white-collar jobs, why accountants? Why not customer service? Or sales? Or project management? Or administrative assistants? Or operations? Or scheduling? Or quality control? Or IT? Or HR? I would argue some of these other areas have more repetitive tasks than accounting. And yet accounting keeps showing up in these research papers. Why? The reason is because accounting is a very lucrative area for software companies, much more than other business areas. Software companies benefit from convincing you to automate accounting. The average ERP accounting system annual fees range between $100k to $200k depending on how big your company is. This is a lot of revenue, and it is consistent revenue because it is difficult for businesses to switch once they sign up. These research papers are a sales pitch. Software companies will say, if you fire your accounting staff at $60k/yr salary each, you will actually be making money by giving software companies $200k/yr instead. What they are not telling you is all the risks you are taking on by replacing humans with a computer. You need to realize that when someone is urging you to take all your money out of the hands of accountants and put it into the hands of software companies that they have their own best interests at heart, not yours. They want your money.
- Accounting is about decision making. This is a concept I have been talking about in every video for the last five weeks on this channel. A business decision is data plus judgement. Computers are really good at processing data, but really bad at judgement. If you think accounting is just processing data, you are completely missing the point of what an accountant does. For example, if you need to make a decision that impacts stock price, are you concerned about stock price in the short-term or the long-term? Because that could result in different decisions. The answer requires judgement that is not possible for a computer. It is not possible because judgement is driven by humanity’s vision for the future. Business is about growth and change. All a computer can do is take the world as it is and try to optimize it. And that is the difference. A good accountant uses judgement to make decisions that improve their business.
Those are three reasons why Artificial Intelligence cannot automate accounting jobs.
- The point of accounting is to watch the money.
- Software engineers have an agenda.
- Accounting is about decision making.
Just to be clear, I am not against automating accounting tasks. Of course, just like in any job, there are many accounting tasks that can be improved through automation. We should always be looking for ways to use computers to improve the effectiveness of accountants. But I disagree with this research paper that claims that all accounting jobs will be replaced with computers. That is just not true. The data shows it is not true. Accounting jobs have actually grown since this paper was published eight years ago.
The real shame about this paper, is that people are making career decisions right now based off this idea that accounting jobs will go away. I am here today to tell you the opposite. More than ever, the world needs good people to become accountants. Society will always need a group of people who are trained to watch the money. If you are watching this YouTube channel, and this interests you, accounting is meaningful work that you can do that can have a positive impact on the world.
Leave a comment down below letting me know what you think!
If you find these videos helpful, please subscribe to my YouTube channel.
Neither Zach De Gregorio or Wolves and Finance shall be liable for any damages related to information in this video. It is recommended you contact a CPA in your area for business advice.