Abacus or AI - It's Still Accounting

It’s 5:00 am and I can’t get this out of my head until I write it down so here I am. If you pop “accounting” in the google engine today’s top result is from Investopedia where they say:

Accounting is the process of recording financial transactions pertaining to a business.

From the beginning in business, owners have needed to understand if their toils were bearing fruit. At the earliest, an abacus was used along with the accountant’s intellect with an aim toward that goal. As time has progressed, so has the quality of tools that accountants have at their disposal to not just collect, but also to categorize, aggregate and analyze. It’s because of this view that, although I am deeply technical, I consider myself an accountant.

Even though I started my accounting career over 40 years ago, thankfully I didn’t have to start my accounting career using an abacus. “Back in the day” my Silicon Valley-based high school had classes where I learned how to use a 10-key adding machine and a typewriter. In my senior year, I even took a technical class labeled “computer-assisted accounting” where I learned the mysteries of punched tape machines and calculated auto part inventory totals with punched cards. A bit later when I finished my accounting degree and was working in the Joint Venture (JV) accounting department of Chevron, I started putting those learnings to work.

In an odd way, a project I did in the JV accounting department 30+ years ago is just as relevant today as it was over 40 years ago. As an A/R clerk, I was tasked with the unenviable job of helping to collect their accounts receivable by providing the documentation necessary after their ambitions for new deals exceeded their systematic capacity to properly deliver it. Then, just as it is now, over time business ownership changed hands. Their case was a much more extreme one because many of their partners were small. Because of that, anytime a partner living their life married, divorced, died or for some other reason divided their interest, new invoices needed to be generated to reflect the proper ownership.

A recent project I completed where we were exporting reports from QBO and calculating ownership earnings showed that even today we struggle with many of the same issues

In my old project, I suppose I could have just sat down and keyed those entries in until someone told me to do different but after studying accounting theory for 4 years, I suppose I didn’t want to settle for a job keypunching numbers. In my mind I did what any good accountant would do - I found a way to improve the process of reporting the financial transactions so it could be done faster and more predictably. Rather than typing out every charge in a spreadsheet and calculating the amount due based on the new ownership interest, I found a way to get a machine to spit out an invoice with the right amount.

In much the same way, today’s transaction feeds from our financial institutions are repurposed using rules so that rather than typing in and classifying every transaction we can simply help write the rules for how that transaction should be classified. Many may not consider it so, but I would dare to say that bank rules are artificial intelligence and learning to leverage them is every bit as much accounting as my early days of learning a paper tape machine.