Do you need a chief artificial intelligence officer?

Register now

In the future, depending on the company size and business type, many organizations will need to consider adding the role of chief artificial intelligence officer. While a full-time dedicated position might seem to be excessive now, the role of a senior leader who is deeply knowledgeable in the broad domain of artificial intelligence is rapidly becoming critical.

The chief AI officer is tasked with providing strategic and tactical guidance and support for exploring AI methodologies. This role would also serve as the evangelist for the process, people and tools that can help achieve real business results with AI or human intelligence augmentation.

Since artificial intelligence, and with it machine learning, is a broad and cross-cutting capability, this individual can focus across the landscape, and research and dig deep into the possible advantages and opportunities to aid business objectives.

Where does the chief AI officer fit in?

Organizations should understand the most appropriate area for inserting the chief AI officer. For example, in organizations where AI and ML can help solve internal optimization or business efficiency problems, the chief AI officer could be an executive within the COO’s group.

Conversely, for those organizations where AI and ML is a capability within their product revenue streams, the role would have a revenue-generating (as opposed to cost optimization) impact and could be placed with the chief product or chief technology team.

There are several good options for where to place this job that will impact the leadership structure of a company. But the best choice depends on the strategic expectations and impacts that AI and ML capabilities will bring to the overall organization.

For smaller organizations, the role might exist as an extension of the data science or analytics organization, using data engineering to promote the development of learning and predictive capabilities.

Don’t believe the hype—Or do

There is significant new hype around AI and ML which can undermine the productive and beneficial uses of these technologies. But artificial intelligence and machine learning are a set of methods in computational science that have been in practice benefiting businesses for many decades.

For organizations exploring this role, review how it has been structured in other organizations and what has been expected (e.g. recommendation, fraud detection, image analysis and so on). Great success has come from the proper implementation of these mathematically-based methods when applied to suitable problems in the business setting.

The real change in recent years has been the focus on data acquisition, cleansing and engineering which is “the food” for a great AI and ML program. The advancements made in data analytics and data acquisition tools, techniques and skills have provided more observations on which algorithms can learn and grow smarter.

This increase in data accessibility and accuracy means we can evaluate and tune our algorithms quicker and with better results. But we still need to understand how systems learn and what problems are best suited to a learning approach.

The key now is for pragmatic application of AI and ML to guide the appropriate and reasonable expectations of AI while continuing to push on the development of new and innovative AI applications, algorithms and supporting tools.

The (many) prerequisites of a chief AI officer

Senior leaders in this role should be very comfortable with statistics, mathematics and data engineering of AI and ML capabilities. They should be able to abstract away from the techniques, methods and tools and see how best to apply AI and ML power from a business perspective. They should be able to build agile teams that can deliver unified approaches.

This person should be a demonstrator at a business level and someone who cares deeply about accuracy and error rates.

As the chief AI officer, this individual should be an evangelist for how to best use AI and ML with all the marketing, development and packaging that goes along with providing cross-cutting, fundamental capabilities into any business environment. They should be able to explain these demonstrations and to clearly outline the limitations of the results without diminishing the enthusiasm for any exciting new capabilities.

This individual should be able to simplify complex topics and influence others with the results tied to business objectives. They should be able to separate fact from marketing fiction, as there can be a confusing array of approaches, vendor products and internal tensions around strategic initiatives.

This role needs to provide a clear, actionable path forward for the chosen AI strategy that allows flexibility but also focuses on realistic delivery along the way.

For reprint and licensing requests for this article, click here.