Fidelity Investments

Artificial Intelligence (AI) is changing the business landscape. Increasing numbers of products and customer experiences are being personalized, service interactions are being guided through intelligent agents, and unstructured data is finally being unleased in novel ways to drive economic value. But it is also driving significant disruption and putting pressure on existing business models. Organizations wishing to compete, must adapt.

To thrive in an increasingly information-driven business environment, new capabilities are required. Across industries an arms race is underway as organizations vie for the skills required to develop, deploy, and manage big-data and artificial intelligence solutions. As a result, many talent strategies are being oriented to in-demand skills like data scientist, data engineer, cloud engineer, solutions architect, etc.

Beyond the technical talent, how does this shift change the talent profile of non-technical business roles like marketing, product, sales, and operations? While there has been a lot of focus on the technical capabilities needed to build AI solutions, the same can’t be said for the skills required by those closest to the customers and business operations that will have to apply and use these solutions. What is needed to make these business professionals “AI-proficient”?

Fortunately, increasing the AI proficiency of business roles does not require everyone to get a PhD in advanced mathematics. However, business and human resources leaders, in partnership with their IT and analytics counterparts, need to think differently about the skill profiles for their organizations, as well as changes to talent management priorities. To get started, organizations should look to augment the knowledge of non-technical business leaders in four areas of understanding: (1) what AI is, (2) how it works, (3) where to use it, and (4) how to manage it.

What is AI?

To be able to identify the types of problems that can be addressed through AI, it is best to start with a realistic and practical description. In basic terms, AI predicts a number or membership in a group. This description is a good one in that it is simple, providing an accurate view of how AI can be leveraged to automate tasks and amplify human talent. Additionally, it helps differentiate the “math from magic” often associated with AI; i.e., AI at this point in time does not rival general intelligence and it’s not about to become HAL form a “2001, A Space Odyssey."

Additionally, it is important to know enough to identify good AI use cases. For example, can a precise goal be set? Is the problem addressable through ananalytic approach that classifies or predicts? Is sufficient data available? Does enough organizational readiness exist to implement an AI solution? How does Artificial Intelligence Work?

It is important to have a baseline understanding of how AI works. This knowledge is important to be able to scope and engage the right resources, understand the capabilities and limitations of the AI solution, and help guide how success is measured.

A strong, working level, understanding of several key concepts is essential, to include:

Supervised versus unsupervised learning

AI, machine learning, and deep learning

Common algorithms, e.g., regression, clustering, decision trees, neural networks, etc.

Common approaches for assessing AI performance

Predominant AI product patterns, e.g., natural language processing, computer vision, recommendation engines, etc.

Where to Use AI?

Shifting from a conceptual to practical understanding, it is imperative to have a broad awareness of current AI applications. Through market intelligence; a body of knowledge should be built around the following questions:

What use cases are being implemented by competitors and across industries?

What use cases are capturing venture investment?

What are the product roadmap priorities for key AI vendors and solution providers?

What insights are the consultants sharing?

How are the use cases being applied across different business functions?

In answering these questions, trends will surface that can generate ideas and guide decisions about where AI can be best utilized within the organization.

How to manage AI?

To lead, or even support, the delivery of AI-based solutions, a solid understanding of the skills, tools, processes, and considerations used to develop and manage AI is necessary. Specifically, deep knowledge is required in the following areas:

End-to-end lifecycle project path for AI solutions

Skill sets required to build AI solutions