Worldwide spending on cognitive and artificial intelligence (AI) systems is forecast to reach $57.6 billion (£42bn) in 2021, according to IDC, with 71 per cent of the CIO 100 expecting to contribute to that investment over the next 2 years. The AI revolution is well underway.

All the major hyper-scale public cloud providers — Google, Amazon, Microsoft, IBM and others — offer AI technologies and toolkits to help customers get going with AI. Meanwhile, end users are exploring a wide range of pilots.

One of the biggest growth areas for AI consists of chatbots. They’re particularly popular with financial services firms such as Capital One Europe, which operates a short message service  (SMS) chatbot to help customers manage their credit card and bank accounts.

More broadly, a lot of mainstream businesses are building chatbots into their website back ends to answer repetitive customer queries; escalating the more complex or higher-value questions to human contact centre agents. This saves time and money, and enhances customer service.

Another AI growth area is natural language processing (NLP). One example of this comes from Mya Systems, which has developed an AI virtual assistant that uses NLP to assist job candidates and employers, through conversational online chat facilities.

Machine learning (ML), a subset of AI, is also proving helpful to businesses across the board. ML can be used to spot trends in vast amounts of big data — both structured and unstructured — and yield actionable insights. It’s being used in a wide variety of ways, from fraud detection to purchase recommendations and up/cross-selling opportunities. In one practical example, Lufthansa is using machine learning to determine ticket prices and flight schedules, and plan its staffing requirements.

Build or buy?

AI can produce transformational results for businesses, but should they build or buy?

The answer lies in whether AI is core to the business, with computer intelligence offering a significant competitive advantage through, for example, improved speed, predictive analytics, innovation or customer service.

With deep learning, which is a subset of machine learning, businesses can intelligently analyse large numbers of images very quickly. This may offer a competitive edge to some businesses, but not to others. So, for some, it may be worth investing in AI and developing it in-house, perhaps adapting the tools offered by the main public cloud providers.

Where AI can help optimise non-core processes, such as customer personalisation, product recommendations, chatbots or cyber security analysis, it could make more sense to use a cloud-based AI-as-a-service or on-premises, shrink-wrapped (end user license agreement accompanying software or hardware placed behind see-through plastic wrapping) or bespoke AI solution.

Either way, you will probably need to add new AI skills, tools and platforms to your IT team to build or manage your new AI apps.

According to online recruitment firm Indeed, the most in-demand AI skills are machine learning, data science, big data, and data mining; and the most popular development languages and platforms are Python, R (an AI-tool), Hadoop, Java, Apache Spark and the AI tools from data analytics vendor SAS.

Most leading cloud vendors are keen to actively engage with businesses to help them develop AI applications. For example, Google recently launched a global competition — the AI Impact Challenge — to help spur the development of applications that would, among other things, help local non-profit innovators make their community — and beyond — an even better place. Public cloud vendors also offer a multiplicity of online resources: technical presentations and videos, case studies and documentation, making their websites a good first port of call.

Further down the line, your company needs to be prepared to consider AI ethics and transparency issues related to human resources and recruitment, safety, medical and healthcare, and personal privacy, among other things.

But, as long as you keep the human aspect in mind when developing your AI, you can attend to these finer points later.

Want to learn more about AI? Read about the state of chatbots – today and beyond. Or find out more about how to stock your organization (or your resume) with AI expertise.