How AI will Change the Accounting Industry

What does the Future hold in store?

---

What will be the role of Artificial Intelligence (AI) in the accounting profession? AI was purposefully designed for boring, mind-numbing tasks; it handles these better than any human, because it works 24 hours per day, doesn’t get tired, doesn’t miss things, and doesn’t accidentally type “5” when it means “6”.

It can do this for any industry, but to be honest, the accounting industry is particularly full of tasks involving repetitive, tedious mental labour. Not only is it hard to enjoy, but it leads to errors, personnel churn, and employee burnout—no matter how great your workers are!

Human beings are immensely complex thinkers who need to be challenged.  They are capable of leaps of logic that machines will probably never be able to duplicate.  While AIs cannot understand our emotion, compassion, or creativity, they can emulate the human analytical process.

The UEA’s accounting firms are looking forward to rolling out Artificial Intelligence solutions, and leading the industry with their innovation.  AI is ready to challenge how we manage traditional accounting practices.  More importantly, accounting people know that AI does not cost jobs; the truth is that AI invariably creates new, better, and more rewarding jobs for people.

Not Better, but Faster

AI machines can assess thousands or millions of possibilities in the same time that a human can assess a single one.  AIs rely on their sheer speed to look at all of the possibilities when presented with a problem.

How do they “know” what to do?  Machines gather “experience” in the form of solving similar problems thousands of times, and then examine the range of answers for a particular problem to see what has worked best in the past.

Of course, they do not have to experience something first hand.  They can use recorded experiences from any source, analyse it, and add it to their own storehouse of information.  A computer that has learned something can share that with another computer, giving it instant experience.

Machine Learning

The famous Big Blue, from IBM, defeated the World Grand Master of Chess, Gary Kasparov, in 1997.  It did so by analysing every available move on the chessboard, and many consequent moves.  It simply knew more than the human.  It won because of pure computing power.

Many years later, IBM’s WATSON computer beat the Jeopardy champion, Ken Jennings.  It was programmed with millions of pieces of data, and training about how to interpret millions of deliberately complex Jeopardy questions.  It had massive computing power and an immense database of handcrafted “experience”.

IBM spent millions of (U.S.) dollars to create the computing power to win at chess.  They spent hundreds of millions on hardware and programming to win at Jeopardy.  Was it worth it?  Absolutely, agree both Kasparov and Jennings.  Humans must explore new frontiers, they insist.  We can’t be afraid of creating greater intelligence than our own.

In the Real World

How limited would your company be if you were afraid to hire anyone smarter than yourself?  You would never be able to accomplish anything more progressive than what you could do on your own.  Smart bosses don’t cower in fear of hiring smarter people.  Top executives know that having smart employees makes their job easier; it enhances the company’s capabilities; it allows them to achieve much more, and with fewer resources.

That remains a big problem in the accounting industry.  We have more and more businesses in need of accounting services that are just starting up, but there are not nearly enough graduates to fill all the available positions.  AI will allow us to do much more, faster, and better, even if we can’t find enough people to hire!

How do machines learn?

Teaching an AI is much easier nowadays.  In a car-assembly plant, for example, with the computer vision systems we now possess, a computer can literally watch an assembly line process.  It doesn’t care if it is a live feed or a recording; to the AI it is all just information.

Analysis shows how every single difficulty which arose was solved.  The AI stores that information as “experience” so when it sees that problem again it already has the answer.  It can combine answers from slightly different problems and arrive at new efficient solutions.

Better yet, that information can be shared with another computer which instantly gains all that “experience” without any time investment.  This is the basis of what is called an Expert System.

Accounting AIs

Imagine if we took all the knowledge and experience of 1,000 expert accountants, and condensed it into a single database.  There would literally be no question that it could not answer about how the numbers related to each other.

Such an Accounting Expert System could recognise a bill arriving, determine if it was real or fraudulent; know instantly who it was for, what it was for, and where it should be posted.  It would also know when to pay that bill, to maximise cash-flow by paying it at the last possible moment, but at the right time to avoid being late.

But it is more than just one bill—it is the thousands of concurrent financial events, in and out, that describe daily business.  Everything would be paid or collected; reminders would be sent for delinquent amounts; inventory would be tracked; sales orders that were delivered, delayed, damaged, returned, cancelled, or refunded would all be recorded. AI would eliminate most of the places were human-error could creep into the equation.

Human Accountants

And what happens to all the people formerly doing these jobs?  They can now deal with special situations outside of the AI’s experience or abilities.  Now they can offer customised, personalised experiences for accounting clients.  Best of all, an already large accounting firm like PushDigits, can handle hundreds more clients, offering them greater accuracy, speed, and lower costs than traditional accountants.

Someone who was formerly transferring a paper copy of a bill into an electronic ledger, running trial balances, and trying to reconcile books from different departments is suddenly available to do more complex, more rewarding tasks!  The AI can handle all of the adding up while humans can now look at creative investment solutions for a client, offering personal services that were previously impossible.

The Future

AIs can do more than just run the numbers. Customers want instant answers, and are perfectly willing to use AI-powered online text ChatBots programmed with all the common answers to questions.  The ChatBots even have the ability to transfer a client to a real human when needed.  With customer-recognition software (knowing names, account numbers, or recognising incoming phone numbers) they can even connect clients to a phone-answerer with a similar personality, giving the client an even better experience.

If ChatBots are insufficient, Voice Synthesis can now provide real-time answers for clients that prefer the telephone. AIs can respond in a human voice that sounds utterly real.  Customers don’t care if it’s not a real person, provided they get the information they are looking for, though the same rules for connecting to a real human still apply.

Better yet, these early-model AIs can now speak in any of the 130 languages that Google Translate™ can transliterate from English.  With modern high-speed data, ChatBots and Synthetic Voices can speak and answer in Arabic or Hindi while the other party speaks Swahili or Russian, or most any other combination.

The Takeaway

Known as “early adopters”—those people that love new technology—accounting firms in the UAE like PushDigits are anxious to lead the way into the AI-powered future.  Artificial Intelligence has been around since the 1950s, but the new and modern version will offer new personalised services, greater speed, increased accuracy, and lower prices, with current, real-time, reliable data.  Let us show you the future of accounting—it’s exactly what you want!