When are we going to look at the human cost of AI?

Another week has passed, and another piece of research has emerged about the potential of AI, but the human cost is starting to become a very big elephant.

The research itself is from SAS and highlights businesses are becoming normalized to the idea of artificial intelligence. The concept of trusting a machine might have been nightmarish in months gone, but now it seems business leaders are interested in the efficiency gains which can be realized.

“Whether we realise it or not, AI has already arrived,” said Peter Pugh-Jones, Head of Technology at SAS UK & Ireland. “From financial services to retail, AI has become more commonplace and we are seeing its use progress from solely back-office support to increasingly front-end, customer-focused roles.

“With the ability to unlock accurate insights from vast amounts of data – in near real-time where required – AI is the key to providing the exceptionally responsive and personalised experiences that customers are demanding and businesses are seeking to deliver.”

According to the research, 30% of companies are not sure if they are ready for the technology, citing a lack of required skills (66%), ROI (55%) and fears of malfunctions (38%), though 37% are planning to adopt the technology within the next two years. These are what would be considered normal statistics; mass market penetration is on the way, but it will be a slow burner.

On the consumer side of things, there is also acceptance. 23% of consumers would be happy to let robots choose, purchase and deliver gifts to their friends and family at Christmas, and 62% of millennials would use automated customer service systems if it provided a better or quicker service. Clearly AI is on the up.

But what about the human cost of automation? This is an area which has scarcely been written about, mainly because the positive messages of job creation in the software segments are getting more traction. This is not necessarily because they are factually more reliable stories, but possibly because there is more PR weight behind them.

Another reason is that people just don’t want to face up to the painful truth that AI will cost a lot of people their jobs and cause a lot of pain. AI will create jobs, but the question is whether it will create the same number as it has made redundant? Another is whether those who have been made redundant are suitable candidates to made use of the new opportunities? Will a data entry employee be an effective business analyst making use of the insight gained from artificial intelligence?

In every industrial revolution there are winners and losers, the difference here is the type of change we are seeing and the speed of change.

Firstly, let look at the introduction of cars into the mainstream as a comparative example. Prior to this introduction, drivers might have operated horse-drawn carriages, but the car made the horse redundant. It did not make the driver redundant. It was a huge societal change and a massive efficiency gain, and of course there was fear. But ultimately jobs were changed, not replaced, and changed in such a manner that the same person could most likely be retrained. Artificial intelligence is taking away the mundane, repetitive roles, and replacing with a highly qualified position.

Secondly, the pace of change is incredibly fast. With previous industrial revolutions, the adoption rates would have been much slower, possibly over the course of a generation, allowing society to adapt. This change is going to happen over a period of 5-6 years. Does this leave enough time to ensure everyone can fit and function as part of the digital economy? We’re not too sure.

Our sister site Light Reading has done a great job at looking through the numbers. Iain Morris highlighted the 20 largest telcos with headquarters in Europe and North America have cut 63,000 jobs over the last 12 months. Part of this is down to acquisition, but part is due to automation.

Deutsche Telekom is one organization which has a vision of managing a network ‘with no human involvement’ through a process of ‘brutal automation’. This doesn’t necessarily sound like it brings benefits to anyone aside from those sitting at the top of the organizational pyramid. But it is not alone. Almost every telco is introducing AI as part of customer services; this will remove jobs from the organization with no or little opportunity to reallocate.

Light Reading has also been speaking to sources which claim AT&T plans to reduce its workforce by a third through the introduction of automation and artificial intelligence. Of course, operators will deny automation is a means to reduce the highest operational cost of the business, salaries, but in an industry which is feeling the pinch of profitability, it might be the case. It certainly wouldn’t be the first time executives have told misleading truths.

This should not be taken as a rant which wants to stop or slow progress. It should simply be seen as a plea to address the unanswered question of what to do with the digital-have nots in society, the ones who are being left behind. Currently the industry is looking at the AI-revolution through rose-tinted glasses, only accepting the truths which pleases it. There is always a downside to change, it just has to be managed effectively. Ignoring something because it is unpleasant, doesn’t make it go away.