After several false starts over the last few decades, artificial intelligence (AI) technology is now seeing real adoption by businesses across multiple sectors. Benefits include increased business efficiency, better customer service and innovative new products and services. We expect usage to follow a similar pattern to cloud adoption, with early adopters pioneering the use of AI, and wider adoption following as the increasing availability of AI tools and services democratises the development of AI applications. In our view, while AI hype is still a factor, a disruptive trend is now well underway and companies that do not embrace it are likely to be at a competitive disadvantage.

CONTENTS
Investment summary: Increasing adoption of AI
An introduction to AI
AI in the enterprise
Learnings
Applications in more detail
How to invest
Appendix

AI is suited to making sense of high volumes of data

As a natural evolution from big data analytics, the use of AI techniques such as machine learning and deep learning bring an element of intelligence to the analytics process. Models can be trained to identify patterns that would take humans too long to find or that they would not think to look for. Areas in which AI can be used include image recognition, trend analysis, prediction, natural language processing, pattern recognition, anomaly detection, personalisation and discovery.
 

Targeted use to augment human capabilities

AI is most effective when used for narrow applications, where there is a predefined goal like fraud detection or language translation. Processes best suited to AI are those with access to large quantities of data or time-consuming, recurring manual tasks that cannot be automated with standard software. Rather than replacing humans, the best models augment human expertise, providing outputs that help inform decisions rather than make them.

Limited listed exposure to pure-play AI in Europe 

Companies broadly fall into two categories: AI-native and AI-enhanced. As is often the case with new technologies, AI-native companies tend to be private, so there is limited opportunity to invest directly in the theme in Europe; listed examples include expert.ai, Darktrace, Insig AI and Mirriad Advertising. Investors may also want to consider companies that are enhancing their products or services with AI as we believe this will give them a competitive edge.

Not without challenges

The use of AI faces several challenges, including the lack of explainability of deep learning models, the risk of bias, the potential for unethical use, the need for access to large quantities of data to train models and the limited supply of data scientists. Regulators are considering ways to address the issues around bias and ethics.