Brought to you by
pandle logo
Share this content

What is Artificial Intelligence and Why It’s Not Something To Be Afraid Of

19th Apr 2019
Brought to you by
pandle logo
Share this content

Artificial Intelligence (AI) has been on the tech radar for long enough that almost everyone is using it during their day at some point. Voice activated search assistants (like Alexa or Siri) are AI. There’s a chance that you’ve engaged with AI at some point, and not even realised.

What is AI?

AI is technology which is designed to behave in a human-like way in response to certain prompts. It can be programmed, or ‘taught’ to look for information and then select an appropriate answer or to present vast amounts of data in an accessible way.

Applications of this range across the ability to ‘read’ and identify the relevant points, return an answer to a defined question, or learn your Netflix viewing habits and to suggest other shows you might like. AI is intended to make things less laborious for humans, performing the heavy lifting of data capture, analysis and conclusion, leaving us more time to get on with other things. Does this mean it has a place in the future of accountancy?

Weak AI versus Strong AI

Weak, also known as narrow, AI is a system which responds with exactly what it has been asked to provide, without actually understanding it. If you ask Siri where the nearest cafe is, it will tell you. A real human will tell you, but also suggest you travel the extra two minutes to the one around the corner, where you’re less likely to contract food poisoning. Weak AI won’t interpret the request, it just wants to give the right answer though it might not always be the best one.

Strong AI has more processing power and is designed to make connections between things, more like a human might. It’s still performing programmed responses, but also tries to predict what else might be needed. Amazon is a powerful example of this. Your search for a hammer will also yield suggestions of what other customers bought at the same time as their hammer, or products associated with it, like some nails. It spots the pattern and includes the data in its response to see if it’s also useful.

What AI can do for any business

For any business, AI can be ‘trained’ to work through tasks and reduce strain on the human workforce. A common example of this is a company website with a live chat feature. Rather than visitors using the site to submit a request through a web form, an AI chatbot could try to help first.

Your website forwards hundreds of requests to your inbox, all asking if you’re taking on new clients. The messages don’t get answered straight away because you’re on holiday. When you get back, you spend the first hour of your day answering each message, but some of them go unanswered because the sender has already gone elsewhere. An AI chatbot could be programmed to recognise the request, confirm that you are taking on the business, and send your new customer information on how to sign up. 

The site visitor often doesn’t even know they’ve spoken to a machine, gets the answer they wanted, and is probably pleased to have had decent customer service at 11 pm on a Sunday. Meanwhile, you can be getting on with the emails too complex for a robot.

What AI can do for an accountant

Accounting services such as auditing can, in some cases, be undertaken by automated processes. With a greater capability for processing and without the need to eat or rest, it can be completed much more quickly. This has value to a client and allows you time to add extra, human, value such as relevant business advice based on the audit findings.

Tasks which are repetitive and don’t require much creative input, such as bookkeeping, will likely grow into mostly automated processes.

Is AI a threat to human accountants?

AI is magnificent at spotting patterns, processing data, and not getting bored or distracted. Tasks which require precision and concentration but have a limited number of responses are perfect to pass on to AI.

But. AI will always require some form of human programming, intervention, and oversight. To revisit Amazon as an example, it can make very useful shopping suggestions based on previous habits, but a human is more likely to realise that now you’ve bought your new loo seat, you don’t need another 20 in different colours.

The algorithms are ever more advanced, but AI has years of development to go before it can take over complicated tasks without supervision. There are some processes that humans will always need other humans for. AI, advanced as it is, lacks emotional intelligence; human empathy for the client relying on you to guide them away from the brink of insolvency gives a much better customer experience.

The growing use of AI in the accountancy sector can create new opportunities which are difficult to ignore, but so far as an accountant’s assistant, not a replacement.