Stephen Martin, head of office at Brewin Dolphin Glasgow, warns against investing in technology that’s been developed for its own sake.

Artificial intelligence (AI) is coming for your job. At least, that’s what some studies would have you believe. Perhaps it’s little wonder then that on Gartner’s ‘hype cycle’ for emerging technologies, we find artificial general intelligence, autonomous mobile robots, and AI PaaS (AI as a platform-as-a-service) steadily making their way towards the peak of inflated expectations.

In truth, the evolution of AI and its effect on the economy are still to be fully understood. Some say it will change the economy as we know it, while others believe its impact will be limited to making our lives and jobs easier. But, between these perspectives, trends are emerging which at least indicate a direction of travel.

For one, it’s now widely anticipated that AI will have a mixed impact on jobs: rules-based, repetitive roles are likely to be replaced, while highly-skilled work will be less affected. In the latter case, it may actually help these workers, by automating large tranches of administration and processing. Then, of course, it will create its own new jobs.

AI will have a social impact too. Societal and political norms may need to be re-shaped to account for the tumultuous changes in the world of work. It may also uncover new possibilities for solving some of society’s biggest problems – but, with that comes some serious risks around relying too heavily on technology. For instance, could the increasing use of machines in the care profession exacerbate the loneliness crisis which is already rife among the elderly?

The same principles apply to people’s everyday lives. AI could become embedded everywhere decisions are made, with smart homes using data to learn our habits and make choices on our behalf. That in itself comes with science fiction-like consequences, with robots anticipating our moves before we make them

Technology for technology’s sake?

It’s a brave new world – but, anyone looking at companies developing these technologies shouldn’t get too carried away. AI makes many things possible, but not all of them will be valuable and only some will have longevity.

Take a look at how the internet has evolved. Online shopping was unheard of only a couple of decades ago – now it’s ubiquitous. By contrast, Google’s smart glasses, which were subject to a lot of industry buzz, proved to be a flop. In fact, if you think back 10 or 15 years, the technology companies being talked about were Ericsson, Nokia, and Oracle. Now, it’s a very different landscape, with few of these businesses featuring the way they used to.

The pace of change has only become quicker in that time and investors in tech companies need to have the gumption to catch a company as it is rising, then know when to sell. In that regard it’s worth bearing in mind that, increasingly, the investable benefits of AI come from data: both the services that harvest it and the platforms which use data to improve their services.

In the case of Amazon, better customer analysis means improved customer targeting and inventory selection, leading to more targeted and successful product suggestions for shoppers. Meanwhile, automation of supply chains reduces lead times, waste, and capital requirements, helping companies to boost profitability. One need only look to Amazon’s milestone valuation earlier this year as proof of such an approach’s efficacy.

Disentangling complex supply chains

Clearly, the value chain for AI-related assets is long and complex. Some of the players, such as Alphabet (Google), feature at almost every stage – first as a proprietary data source and finally into applications, such as those found in self-driving vehicles. Those businesses are well hedged to realise any benefits at all stages of the typical cycle.

But, the workhorses of AI are chips. They are at the heart of the data centres where AI training and the use of that learning take place. Some of the big chipmakers, such as Intel, work with AI companies from the design and production phases. But, elsewhere the process is being separated, with firms such as ARM specialising in design while outsourcing the actual manufacturing.

To sidestep the risks these complexities may pose, the best play could be to look at ancillary industries. When it comes to robotic cars, for example, that may well mean considering tyre companies. They will benefit from the expected increase in miles driven, regardless of which technology provider wins the race to perfect the technology.

Capital gains

Nevertheless, the real winners will be the people at the top of the AI revolution: the owners of capital. They stand to gain the most through returns on investments, while the economic gains should increase the amount potentially available to fund a social safety net (i.e. a universal basic income), if one is required.

Indeed, rising living standards are unlikely to flatten the wealth and income inequality we’ve seen increase in recent years. But, this could create opportunities for brands to target different segments of their markets, facilitated by logistics, platform and data companies, as well as property in the right places.

In the past, brands held the advantage by monopolising distribution, with shelf space at most major shops – think Unilever and Colgate-Palmolive. In the future, companies with the greatest intellectual property are likely to have the edge – LVMH, Apple, and L’Oreal, for example – using it to insulate themselves from competitors and maintain a lead in technological development.

AI opens up a range of possibilities – for society, for businesses, and for investors. But, like most things in life, there will be winners and losers in each of those categories. For investors, thinking carefully about the technology they’re backing, taking the three points above into account, could help them root out the companies offering real value through their products, rather than just a fleetingly popular new trick.