By John P. Desmond, AI Trends Editor
Companies are taking AI training into their own hands, hiring outside firms to help their employees to learn about AI and often picking up the expense.
Training is a big market. The annual North American workplace training market is estimated to be $169 billion, according to an estimate on Statista. Spending on annual workplace training averaged $83 billion from 2012 to 2019. The share of US companies that partially or fully outsource training is 53%.
Royal Dutch Shell could be a model. The company is expanding an online program that teaches AI skills as part of an effort to cut costs, improve business processes, and generate revenue, according to a recent account in WSJ Pro. Of its 82,000 employees, about 2,000 have expressed interest or been approached by management about taking AI courses through Udacity, the online education company. These include petroleum engineers, chemists, and geophysicists.
The courses are voluntary and employees can complete them at their own pace during work hours; the company covers the cost of the training.
“Artificial intelligence enables us to process the vast quantity of data across our businesses to generate new insights, which can keep us ahead of the competition,” said Yuri Sebregts, Shell’s chief technology officer, stated in an email to the Wall Street Journal.
In a pilot program with Udacity in 2019, Shell trained 250 data scientists and software engineers in areas including reinforcement learning, a branch of machine learning where algorithms improve by trial and error. The scientists are applying the models to better predict machine failures, reduce carbon emissions and process seismic and geological rock formation data. This could assist in decisions about where to drill.
“Technology is moving so quickly that if you’re not continually training your people, you’re going to get out of date,” stated Dan Jeavons, Shell’s general manager of data science.
The trend of companies taking AI training inside is a boon for Udacity. Courses normally cost $400/month for individuals. Udacity Chief Executive Babe Dalporto was quoted as saying that the firm’s courses for corporate employees are likely to be its largest business this year, with interest accelerating over the past two years.
“Any Fortune 500 company is realizing that AI is going to be disruptive,” stated Sebastian Thrun, Udacity’s founder, president and executive chairman.
IBM Launches AI Skills Academy
In December 2018, IBM launched its AI Skills Academy (AISA) with two objectives, according to a recent account in Fortune. First, it was to teach employees how to integrate AI into their own jobs in the company, ranging from marketing to supply chain responsibilities. Second, AISA teaches how to collaborate with clients to use AI in their businesses. The program has two tracks: technical and non-technical, with four levels in each track, basic to expert.
Some 2,200 IBMers had started the training in its first six months or so; IBM expected 4,000 to complete all four levels by the end of 2019. “That’s just for openers,” stated Obed Louissant, VP Talent, Watson Health and Employee Experience, IBM. “All of our employees will eventually be trained in AI.” New courses are continually added, for instance in general management.
While it might have been risky for IBM to commit to training its own workforce in AI, for fear of making them more attractive to other employers, attrition from the trained group has been lower than average so far. In surveys on what motivates them, training came up big. “They are most interested in keeping up with the cutting edge in technology and continually learning new skills,” stated Louissant. “So offering them new training is a retention strategy.”
Learning and Development Professionals Tuning into AI
Learning and development professionals need to stay on top of the impact of AI on their market, suggests a recent article posted on the blog of Virtual Speech, written by company co-founder, Dom Barnard. For example, teaching styles need to be personalized and customized for males and females. AI can be employed to help focus on weaker areas of the learner, recommend suitable content, predict needs based on the learner’s role, and auto-generate content using content creation algorithms.
Learning preferences can range from video tutorials, written content, in-person training, gamification, and audio-guided presentations. An AI-powered learning management system can adapt, offering video tutorials to certain employees, and auto-transcribing the videos to text-based articles for other employees. The system could create visuals based on written content, and suggest the employee take an in-person training day if they struggle with a certain section.
Online assessments can also adapt, tailored to the individual’s ability and progression. An example is Iris, developed by PluralSight, a technical training provider. Iris adjusts question difficulty and skill ratings as the learner progresses. Using natural language processing and machine learning, it recommends assessment content.
Virtual Speech combines employees training with virtual reality to give users a realistic way to practice different soft skills and provide instant feedback on performance. The company offers a white paper on VR and AI for Soft Skills Training.
Savvy employers can use AI to help “upskill” employees, to be proactive in ensuring the most relevant training tools and knowledge content is available to employees when they want to learn. Private industry is of course responding. An example is startup Guider of the UK, which uses AI to match employees with mentors and to manage the mentoring process.
AI Can Help Integrate Training into the Routine Workflow
AI can help integrate workplace training into the regular workflow, suggests a recent entry on the blog of WhatFix, a software company whose platform helps users create interactive walkthroughs for learning to use Web applications and other software.
An AI-enhanced learning management system can personalize training paths for each individual; improve completion rates; build content at scale; frees up learning and development staff for more in-person training sessions; remove training bias; and measure effectiveness.
The system can also venture into new learning models such as game-based courses and storyline-based games. The behaviors are tracked to customize learning to make it more interesting for employees.