The State of Artificial Intelligence in 2019: Are Businesses Buying the AI Hype?

As of January 2019, just 14% of large businesses have successfully deployed an AI initiative. However, by the end of 2021, that is expected to have increased to more than 50%, with another 29% to follow suit in by 2022. Interest in AI has increased seven-fold since 2012 as a result of frequent coverage of the cost and time-saving benefits of successful AI applications.

Most businesses are now aware of the technology’s potential and how it can be applied to their respective industries. The insurance, IT, telecoms, retail, health, and media sectors have been particularly quick to respond.

More than 40% of businesses in those fields have already deployed AI or plan to within the next 12 months. Yet, the government, education and charity sectors are falling behind, with less than 30% of organisations expecting to deploy AI by 2020.

AI’s early adopters are finding a wide range of uses for the technology. According to Gartner, 10% of enterprises claim to be using ten or more AI applications in 2019, despite 72% admitting that they had yet to successfully deploy AI just 12 months earlier.

Chatbots, AI systems that can simulate human conversation, and process optimisation, AI systems that can improve the efficiency of business operations, were the two most common uses. However, enterprises continue to find innovative ways to make use of AI, from screening job applications to detecting fraud.

Investment in AI technology continues to increase, but data shows that a gulf is emerging between those that invested early in the technology and those that didn’t.

In the past year, 69% of early adopters have decided to increase their investment in AI. Those that are interested in or currently developing AI projects are also increasing investment. However, AI laggards – those that have been slow to consider the technology – continue to put off investing. Overall, the portion of companies that have implemented an AI strategy has grown, but for those that fall under the digital laggards category, this figure remains unchanged.

For those that are implementing AI, most are choosing to outsource development, with 56% admitting that they buy AI solutions or outsource development to third parties. That is compared to just 36% who say they build AI solutions in-house. This could point to a lack of available talent for businesses to hire, with 58% of all AI professionals currently being snapped up by the technology and financial services industries.

Despite fears that AI and automation will lead to widespread job loss, studies have found that AI is creating an increasing number of new jobs too. AI-related job postings, compared to the total number of job postings on the job listing platform Indeed, have more than doubled in the past 18 months.

However, while there is plenty of work going in the field, data scientists feel that they are being held back by a lack of talent. For every one suitable candidate, there is 2.3 AI-related jobs available. Demand is expected to significantly outweigh supply for some time.

It isn’t all doom and gloom for AI professionals though – the lack of talent continues to fuel high salaries in the field, with the average AI engineer earning $224,000 annually working for companies like Uber, WalmartLabs, Netflix, Facebook, Salesforce and Google.

Businesses may be struggling to attract AI talent, but finding AI solutions shouldn’t prove too challenging. One in 12 new startups founded in Europe now considers itself to be an AI company, and that figure has been continuously climbing for the past six years. With venture capital investment having also seen significant growth, both in terms of number of deals and funding received, since 2016, more and more startups will appear hoping to capitalise on the emerging market.

With businesses looking for ready-made solutions, it is unsurprising that 87% of AI startups consider business-to-business as their core market, with health and wellbeing, finance, media & entertainment and retail among the most served industries, likely due to the rich data sets that they already hold.