Closing the AI and Machine Learning Skills Gap: What Boards Need to Know

Implementing artificial intelligence (AI) and machine learning (ML) throughout your organization requires a collaborative, interdisciplinary approach between data scientists, statisticians, machine learning engineers, AI researchers, app developers, and line-of-business experts. According to LinkedIn Corp.’s 2018 Emerging Jobs Report, 6 of the current top 15 emerging jobs are related to AI and ML. However, 69 percent of executives in the US mentioned that they are facing a moderate to extreme skills gap among their employees. At the same time, boards and management also need to be trained on machine learning and AI concepts so they can identify relevant business opportunities and understand how to oversee the implications of these technologies on their stakeholders. So how should boards think about identifying and closing the skills gap?

Identify areas where AI and ML will play a key role. AI and ML initiatives are most successful when business imperatives are well understood. Boards need to ensure that management—or someone that management trusts who acts as an advisor—has a strong foundation in the applications of AI and ML, the recent advancements in the space, and how other organizations in their respective industries are using AI and ML to drive their businesses forward. Management must be strategic about the projects they invest time and resources in, and be ready to report on outcomes over time to the board.

Evaluate pertinent AI and ML skills. AI and machine learning encompass a broad variety of use cases, techniques, algorithms, and frameworks. Mapping business needs to the areas where AI and ML can have the maximum impact is the critical first step in narrowing your organization’s areas of focus, and boards should ask to review how this focus was determined. Common subsets of AI and ML include natural language processing (be it speech or text), machine vision for image understanding, time series prediction for forecasting, classification problems, personalization and recommendation engines, and anomaly detection for fraud prevention, among others. Often, AI and ML applications intersect with other related areas such as the Internet of Things (IoT) and Robotics—for example, when your organization is looking at predictive maintenance solutions, they need to combine time-series data from the IoT sensors, and apply machine learning on that data to make timely predictions. Another area of consideration for management should be to identify the ML techniques your organization will require across deep learning, neural networks, reinforcement learning, unsupervised learning, and other spaces.

Understand how the skills in your organization are mapped. Management with a strong foundation in AI and ML can more effectively identify the key skills needed over a longer time horizon. Doing a thorough analysis of the skills required and the current set of skills in your organization is a critical role to be performed by management and the board should ensure that it is done. Since successful AI and ML implementation requires an interdisciplinary approach, management needs to map the skills needed across data scientists, ML specialists, application developers, statisticians, and other subject matter experts such as economists, medical professionals, and more. Once the gaps have been identified, management needs to build a strategy that includes both attracting new talent as well as upskilling your current workforce.

Look into how your workforce is being trained and upskilled. AI and ML skills are becoming increasingly more accessible—not everyone needs to have a PhD or be a researcher to add value to your initiatives. Investing in an upskilling strategy will not only help close the skills gap, but also ensure that you retain key talent. Boards should ensure that management leverages existing curriculums and tailors them to your needs. Moreover, pre-trained AI services provide ready-made intelligence for business applications and workflows. AI services easily integrate with business applications to address common use cases such as personalizing recommendations, modernizing your contact center, improving safety and security, and increasing customer engagement. Most of these services don’t require deep machine learning expertise, and developers can easily get started.

Make sure that learning is fun and social. While your board may not be getting in on the action, it can set the expectation that learning be made into a social event that employees want to be part of. Organizing bootcamps, hackathons, and learning events can accelerate the learning in your organization. Increasing access to learning tools like DeepRacer and DeepLens puts machine learning in the hands of every developer, allowing non-ML experts to learn and build on their own. Organizations like Morningstar are using DeepRacer challenges to harnesses their employees’ enthusiasm for machine learning and artificial intelligence. James Rhodes, chief technology officer at Morningstar, says, “It provides hands-on training across the company and accelerates our practical application of machine learning across our investing products, services and processes.”

Ask management to build a center of excellence. Training platforms like Udacity and Coursera offer advanced machine learning curriculum and scholarship challenges. Amazon’s ML University offers a variety of free courses for business decision makers to developers and machine learning experts. Boards should encourage management to leverage these great learning resources and build their own customized curriculum based on these goals. Identifying a few key AI and ML champions within your management team and establishing a center of excellence will help build a virtuous talent cycle through continuous learning.

Seek plans for how consumers of the technology will be trained. Finally, it’s also important for boards to keep in mind that AI and ML can have a transformative impact on your tools and business processes. Boards must think about the long-term effects of implementation early on. Does your company need to train your customer service agents, your hospital’s medical staff, your security engineers, your maintenance staff, or more?

Learn more about building a comprehensive AI strategy for your organizations.