How to avoid the pitfalls of artificial intelligence

Technology is wonderful – when it works! With all the hype around artificial intelligence (AI), most industries are working hard to find the solutions without the benefit of the lessons of history to guide them.

As with any innovation process forging into the unknown, trial and error is inevitable as we fail forward into the future.

After getting really annoyed with some bright spark ideas gone wrong, I became intrigued by the concept of artificial unintelligence and what drives it.

What makes AI different from previous technological innovations is its ability to learn independently from the data – the quality of the thinking that goes into that coding is vital!

In business the purpose of AI is to provide an optimal customer experience, while saving cost and human resources.

Three questions that emerge are – what are the assumptions we are making around the role of AI in our business, how consciously are we interrogating the unintended consequences we are coding into our technology and what do we need to incorporate into our decisionmaking to ensure the best use of AI?

The phone rings in the early evening from a number I don’t recognise.

It was not picked up by my SpamCaller App and a BOT (software application that runs automated tasks) tries to sell me an insurance package. My response was to disconnect the call and block the number.

Now I have no issue with BOTS – I engage with them often online and find them very useful.

They are especially useful when there are scalable options to interact with humans through online chat facilities or calls to the call centre.

For decision-makers in the selling process the assumption that you can take a script used by a human and automate it, is flawed.

The lure of cost savings with this type of technology is strong, particularly because human contact is expensive.

The unintended consequence is not only the loss of a sale, but also the damage to your customer relationship, customer retention and brand promise.

Rich video content, with its ability to personalise and create emotional connection, is now the norm.

Commercialised by platforms like YouTube, they now force viewers to watch ads rather than giving them the option to view or skip the advert.

From a business perspective it poses a conundrum for advertisers on how to engage with their customers so they choose to watch the advert.

As AI moves into the services industry, we will see more automated decisionmaking in professional and medical services.

It is critical that leaders expand their thinking and decision-making across silos to holistically evaluate the full impact of the technology on the customer and the business.

Any AI rollout should have rigorous analytics and machine learning built in to evaluate quickly how these approaches are being received, the influence on bottom-line results and the longterm relationship with customers.

Is our collective enthusiasm to apply cost-saving technology to every aspect of business resulting in bad system design, which will inevitably have far-reaching unintended consequences?

A great technique to mitigate this is the zoom-in, zoom-out approach.

In your strategy planning and implementation, zoom into all the detail and then continuously zoom out to evaluate effect, risk and experience.

AI is there to make business better – so let us do it intelligently.