Get in touch with us. We'd love to hear from you.
Addressing the Skills Gap with AI
Over the last few years, conversations about automation have shifted focus from the elimination of jobs to augmentation. As the world moves to further integrate AI and machine learning tools into daily work, companies are also moving to update the workforce. According to LinkedIn Learning, the identification and assessment of skill gaps is currently the largest focus area for talent developers. Skill gaps in general account for four of the top seven focus areas for 2019.
Technological skills—including skills in AI and machine learning—are the most in demand, but they aren’t the only areas affected by the AI revolution. In 2016, 92% of executives called soft skills as important or more important than hard skills. Consider a healthcare worker whose job has been ‘augmented.’ AI-powered software can help her complete complex paperwork. With less time spent on administrative tasks, she can focus on patient care. Where attention to detail and administrative competency were once differentiators in her position, now interpersonal skills are. This new era of automation restructures skill requirements, even for positions that don’t require any knowledge of AI.
The story is the same across every industry. AI-driven research platforms can help lawyers sift through case files, reducing billable hours. Call centers can automate simple requests, reserving human operators for escalations. Changing skills requirements can improve our processes, but they also create mismatches between workers and their current jobs. The changes have implications across the entire labor market. For companies to stay competitive, talent developers must embrace continuous learning.
But the AI integration that drives these skill gaps can help bridge them. With the right data, machine learning can help businesses adapt to emergent skill gaps, evaluate talent and even retrain at scale.
Measuring the Gap
A 2012 study from Harvard Business Review suggested that the shelf-life of a bachelor’s degree is only five years. Job requirements have substantially altered since that study was written. This means that our employees require continuous growth in order to remain proficient at the same job, but it also highlights an important challenge for learning organizations: the future that we are preparing for isn’t permanent. Reskilling and upskilling are not one-time events. In order to create an adaptive workforce, we need a culture of continuous learning.
Perhaps the greatest caveat in AI implementation is its need for a large amount of clean data and metadata. Skills themselves need to be assessed and described by humans before those descriptions can be parsed by a machine. In many organizations, that data exists, but it isn’t integrated. It exists in performance reviews, job descriptions and professional learning results. It’s implied in internal social networks and explicit in competency frameworks. AI tools can use natural language processing to crawl these systems in order to find a skills baseline. As more skills information is incorporated into the system, the dataset grows, allowing for greater granularity and greater power of inference. KPIs between higher- and lower-performing employees can be compared. With a robust enough system, that information can be used to formulate a development plan.
Bridging the Gap
The primary role of machine learning for learning and development (L&D) is to augment the relationship between the learner and the content. With that in mind, there are a few approaches for the incorporation of AI into the training process.
Algorithms can crunch employee performance data and match it against an organization’s predetermined skill gaps. Custom assessments and employee learning history can be used to establish an individual skill baseline. This combination can be used to set out a personalized development path for each employee. As more employees use the learning platform, it becomes better at predicting the needs and interests of its users, similar to how YouTube recommends relevant content based on the results of similar users. The path of development can be further customized by ability (according to assessment performance) and preferred media (according to user history).
The above generally presumes the style of eLearning courses that are standard now. These courses are an important part of continuing professional education, but some more contemporary methods of learning also benefit from AI tools. Chatbots can use natural language processing to offer one-on-one coaching on a variety of topics. Just-in-time learning, which is learning that doesn’t occur until you need it (like a primer on presentation technique), can be offered based on location, such as at a client’s office or a conference. The goal of these technologies is to build a culture of learning that is continuous, adaptive, timely and targeted. It builds the kind of supportive learning environment necessary for building and developing top talent.
Beyond the Gap
Comprehensive learning strategies are not simple to develop. Training does not happen for free. And while many organizations have increased their learning budget in recent years, others still insist on hiring—rather than reskilling—as the solution to the skills gap. Available data suggests that this is short-sighted.
Research from the Society for Human Resource Management (SHRM) puts the immediate cost of replacing an employee at 50‒75% of their annual salary. Final costs are estimated between 90‒200%. Among millennials who plan to leave their current company in the next two years, 71% cite a lack of training opportunities. Ninety-four percent of employees would stay at a company longer if it invested in their career. Developing your employees doesn’t just keep them competent. It keeps them around. And it’s easier to reskill a high-performing culture fit than to train and acclimate a new employee. The message is clear, and it’s relatively dire. If we don’t reskill our workers, we will continuously pay to replace them.