Which sectors will be hit hardest by AI?

Gartner analyst Anthony Mullen tells CRN which industries will be most impacted by artificial intelligence

Gartner analyst Anthony Mullen speaks to Marian McHugh about the influence AI will have on multiple sectors in the next decade

A recent report from PwC indicated that artificial intelligence (AI) will create as many jobs as it will destroy over the next two decades. While the news was optimistic about the technology's effects on workforces, it brought into stark relief those jobs that will become obsolete as AI becomes more intrinsic to businesses.

Anthony Mullen, a Gartner research director specialising in AI technologies, told CRN that the roles most at risk of being lost to AI are those that are focused solely on one task or "domain of knowledge".

"The jobs that will tend to stay in the hands of humans in any industry will be those that are multi-disciplinary; [those jobs] where people are bringing design skills, software skills and people skills into the mix," he said.

Earlier this year, Gartner conducted its own survey regarding AI perceptions, plans and implementation. It surveyed 848 respondents working in enterprise across North America, the UK, China and India. The respondents had to be employed by a company that was either piloting or deploying at least one of the following AI techniques: natural language processing, computer vision, AI physical robots, process and decision augmentation.

Mullen said each industry has its own "slant" with regards to how it is implanting the technology and stated that AI will affect all industries in the next five years. He picked out the five sectors that he predicts will see the most changes in the next few years courtesy of AI.

Finance

Financial services firms are among the biggest investors in automation, according to Mullen. He said companies in this sector are doing lots of pilots of prototypes, as well as "envisaging" a digital workforce alongside a human one.

"Most organisations are not likely to see net new business [as a result of AI] for the next four to five years, so most are now looking to reduce operational costs, so better risk models, better loans around repayments.

"Most medium-sized to large banks worldwide are already making big changes in their contact centre and their internal IT support desks, and beginning to automate some of the more routine service requests," stated Mullen, adding that the implementation of virtual assistants by these enterprise companies can lead to "really huge" savings.

Transportation

Mullen said Gartner's survey indicated that workers in transportation are among the most worried about the impact of AI on their jobs. The PwC report pointed out that this will be one of the sectors most affected, with 22 per cent of net jobs expected to be lost by 2037.

Mullen reported that services in the transportation industry are already seeing signs of impact, citing the example of Japan, where many train stations now use robotic assistants to help customers.

"The visible face of transport to lay folk is the example of driverless trains. [It is also] using computer vision and analytics models for planning and routing," he said.

Healthcare

The analyst claims that healthcare provides an opportunity for AI to bring supply and demand to pressurised health services, citing the use of AI to diagnose dermatological complaints and analyse X-rays as examples of where AI "already outperforms" its human counterparts.

"In the UK, the 111 NHS phone line is doing some experiments with chatbots. The government can certainly play a part here in terms of investment," Mullen added.

"We shouldn't forget that the major smartphone manufacturing players - Apple, Google and Amazon - all have partnerships with healthcare providers too and that is all designed to reduce the load on national systems, to help people be a little bit more autonomous," he said.

Mullen predicts that the shift in support to AI technologies and the massive problem with the supply of services and demands of patients will push more effective tools out to patients and thin the traffic that is going to national services.

Manufacturing

The PwC report claimed that the next 20 years could see the manufacturing sector lose a quarter of its jobs. The Gartner survey lends credence to this, with 56 per cent of respondents indicating that AI improves the workflow process and inventory in manufacturing.

"A lot of the computer vision technologies are reinventing building information management systems, allowing them to speed up the design process as well. By using things such as predictive analytics, they are much more able to predict when machines or equipment are likely to fail, so the replacement and repair is much more efficient as a result," the analyst explained.

He also said that he is seeing an increase in the trade and barter of materials being handled by algorithms rather than people.

Customer services

Gartner conducted a similar survey with consumers at the same time as it surveyed enterprises. One question asked respondents, if they were faced with choosing an equally capable human or an AI to undertake a specific job, which they would opt for. The result reported that 38 per cent of people would choose AI to handle customer service, with 16 per cent indifferent.

More than half of these people would be either happy or OK with you having an AI support them in customer service, so the demand for this is there," Mullen said.

"I characterise this as the great training period. A lot of the low-hanging fruit or the easier customer service questions can be dealt with by a bot now and if the bot can't do it, it hands off to a human and every time the human answers the question, that goes back into the training data. We're going to handle all these exceptions and oddities over the next few years to the point where a much larger percentage of customer service tasks will be handled by a bot."

In the next three to four years, routine customer service queries will be handled by bots, Mullen predicted. He also said that by the end of next year most medium-sized and large organisations will be using some form of virtual assistant in a customer service capacity.

However, the analyst warned that businesses cannot go digital without fully embracing AI technology.

"It's one thing to say every business should go digital, but sometimes the existing business and processes are just digitised - it doesn't necessarily mean you have the best, most elegant approach to process flows or services. I see AI really helping businesses be way much efficient [in this regard]."