Fear or future? Getting to grips with the AI skills gap

Generative AI – tools such as ChatGPT – has bulldozered its way into our work and personal lives. How can HR and learning teams upskill employees to use it productively and overcome the concerns over its potential to transform, or even replace, jobs?

There’s a lot of fear around the impact artificial intelligence will have on the labour market. One of the most dramatic recent predictions came from Goldman Sachs, which suggested that generative AI – meaning tools such as (but not limited to) ChatGPT – could replace 200 million jobs.

At the same time, HR teams are rapidly reviewing policies and processes, recognising that employees’ use of AI in their everyday roles could require some “guardrails” to ensure they are using it fairly and ethically. In March, research from Gartner found that half of HR professionals were introducing policies to regulate the use of AI at work.

But in the rush to place boundaries around how employees use AI tools, could HR be fuelling the fear and miss out on the benefits?

Role transformation

According to learning software company Skillsoft, there is a “fast-growing AI skills gap” that companies need to address. Its research has found that 96% of L&D teams believe that AI will transform job roles, yet 63% of organisations in EMEA have not provided employees with any AI training in the last year.

Only 58% of respondents said their organisation’s investment in AI skills building would be enough to close the gap.

The company has recently launched an ‘Intro to ChatGPT’ course that educates learners on how generative AI works, its risks and limitations, and how to create effective prompts when using the tool.

Mark Onisk, Skillsoft’s chief information officer, says there is still a large proportion of organisations that are taking a wait-and-see approach with generative AI.

“There are many who have placed a moratorium on this until it gets mainstreamed,” he says. “They’re saying they don’t want to put themselves at risk. But those who are looking at it are rapidly skilling up, starting with the basics of what the technology can do right up to the ethical considerations and the guardrails they need to put in for compliance.”

Onisk argues that there are three key considerations when integrating generative AI into the workforce: the technical skills needed to use it in someone’s role; the leadership and professional skills required to use it effectively; and how you manage the potential risks and compliance issues associated with it.

“This should begin with the foundations of what generative AI actually is,” he adds. “It’s important not to get too indexed on the tool itself, there is much more out there than ChatGPT. Once employees understand that you can move on to policies and responsible use.”

Getting ahead

If organisations are taking an overly cautious approach, they may find that their employees and customers are already experimenting with generative AI, and some may bring it into the workplace before policies or training are in place.

Potential new employees are already using the technology before they even join, according to research from CV-Library. More than four in 10 are using ChatGPT to help write their CV, 37% to write their cover letter, and 35.2% to get a better understanding of the job they’re applying for.

Lee Biggins, founder and CEO of CV-Library, says this is “both expected and understandable”, and that the technology can help candidates create a more powerful CV so they stand out in the hiring process. For HR, it’s a signal that they need to equip themselves and hiring managers to understand the benefits AI can bring as well as any red flags that might be raised.

In turn, learning and development teams need to consider adding or developing training content that builds those skills – from practical uses of ChatGPT such as generating lists, to the ‘softer’ skills greater use of AI necessitates. Skillsoft’s programme includes modules on ethical issues around AI, for example, such as how far we should trust its outputs and the importance of casting a human, critical eye over AI-generated work.

Job design

As well as developing learning paths that support the transition to a more AI-friendly workplace, businesses need to think more broadly about how jobs are designed and what that means for individuals’ career progression.

Furniture retailer Ikea has shown how this can be done, reskilling its call centre workers in interior design as AI ‘bots’ take over an increasing number of straightforward queries.

One of the key skills deficits noted by Skillsoft among employees was ‘soft’ skills such as leadership (54% of survey respondents) and communication and problem solving (52%).

Onisk adds: “It’s about how we reimagine work, how business processes look in the age of generative AI, and how leaders support teams through that.

“In every organisation, there will be a threshold where you can’t leave the AI to run the business – some things require human intervention. If you’re a salesperson, for example, you can automate a list of common objections but if there’s a pricing conflict, a human needs to step in.

“Once you understand that, what skills transformation is needed? How do you lead people with empathy through disruption? There are low-stakes roles that can legitimately be automated, but that creates a lot of anxiety.”

Learning platform Degreed has been creating career pathways for its own employees. “The CEO has put us all on an upskilling journey,” says Janice Burns, chief career experience officer at the company. Degreed gives employees a $1,200 education stipend every year and staff are encouraged to start learning about Chat GPT and generative AI.

In weekly team meetings, presenters can “show and tell” what they’ve learned about the tools that week, and what they found tricky. The company is in the process of introducing generative AI into some internal tasks and processes so they can be more efficient. “We’re learning together – no one feels like they’re the one that doesn’t know what they’re doing,” she adds.

Creative support

Digital agency Evo has been using AI to perform tasks such as searching for imagery to go on a new site for a client.

“We can lose hours searching for the right images to go on a new site. With some cleverly crafted prompts, Midjourney [an AI image generation tool] can pump out high-quality options that are exactly what we’re looking for – no big-budget photo shoot needed,” explains marketing manager Helen Hawkins.

Another tool is Figma, an AI integration that will further enhance and assist the design process, she adds, while the agency uses Google Ads’ AI tools daily to speed up account management and search engine optimisation for clients. “Machine learning creates predictive data insights which further enhance the results. And that’s us just scratching the surface,” she adds.

Evo is embracing, rather than fearing, the impact of AI. “As a digital agency, you’d expect us to be all over ChatGPT and other AI tools and we are – it goes with the territory. AI is something that excites us rather than freaks us out,” Hawkins says.

“For us, supporting the team to use as many AI-related tools as they can to improve processes is a given. We’re all tech-minded and curious so playing around with these tools is encouraged and naturally part of the job.

“More importantly, we’re focusing on upskilling the web developers and designers of the future to it. For them, it will be the norm but education is lagging well behind in offering the kind of training they’ll need to enter this world. A world that changes at lightning speed. That’s where we think we can help the most.”

The company is also supporting work experience students to get to grips with ChatGPT. They have been using the tools to help them create their own sports brand and website.

“We’ve had them follow the web design process from the product to the branding to the full web design. Part of that process for us now includes using ChatGPT to come up with metaphors which in turn form part of a Midjourney prompt to generate images,” she adds.

“Whereas previously we would have been limited to using generic stock imagery for these tasks, Midjourney allowed them to create specific products and promotional imagery focused on exactly what they wanted. These tools are getting more powerful all the time, so it’s important the next generation is aware of them and learn to use them in the right way.”

Human element

While generative AI will take over much of the transactional work and save time, employees still need to be able to lead teams, manage projects or deal with humps in the road – the “human” aspect of jobs that AI can’t replace.

Neil Finegan, leadership and management tutor at DLC Training, says it’s vital that companies prioritise soft skills training as their use of AI increases.

“Adopting a learning mindset is essential to get the most of AI tools and to stay up to date with their limitations, capabilities and the broader impact they may have in the workplace,” he says.

“Emotional intelligence is also not to be overlooked, particularly as AI makes its way into the professional environment. AI tools lack human understanding and empathy that are an essential component of HR. Skills such as active listening, understanding non-verbal cues, conflict resolution and maintaining positive relationships are crucial.

“Finally, critical thinking is arguably one of the most important non-technical skills that should be prioritised. Using AI is a great starting point for task automation but they should always be viewed through a critical lens, particularly with ethical considerations in mind.”

Julia Morgan, director of permanent recruitment at IT resourcing company Experis, agrees that human oversight will remain essential.

“Human soft analytical skills, like critical thinking and contextual understanding, are vital in ensuring accurate and meaningful insights. With AI handling more mundane tasks, people will have more time and energy to dedicate to other areas, such as strategic thinking, collaboration, and exploring new opportunities,” she says.

“This shift allows individuals to focus on their unique human strengths and drive progress. Furthermore, AI can fuel creativity by answering hypothetical questions, providing new perspectives and insights, and opening possibilities for innovation and problem-solving.”

Ethical approach

Furthermore, it’s important to take an ethical approach to their use. A survey by the website B2B Reviews recently found that one in 10 HR employees at technology companies are using ChatGPT to write employee termination letters, for example.

Another poll from Aimpoint Digital revealed that one in 10 UK employees had used AI at work without their manager’s knowledge.

What we don’t know is whether those letters were accompanied by human, wraparound support, potentially making their source less relevant (after all, many HR professionals use template documents for just such occasions).

“Like many other sectors, those in HR are feeling the pinch when it comes to time and cost-saving measures which means it can be tempting to fall back on AI to help us in our day-to-day,” Finegan adds.

“While neither inherently good nor bad, it’s important to remember that AI lacks the most important element of the HR profession – humans. It’s vital that if you do decide to start incorporating AI into your role, you do so responsibly, ethically and within reason.”

ChatGPT can be a good starting point for something like a dismissal letter, he advises, but HR teams must review the output carefully as “AI generated text can often lack depth as these tools have a limited understanding of the nuances of natural language”, meaning they can end up being “cold and robotic”. Furthermore, adding in individual details can help to personalise documents that are initially generated by AI.

Part of the learning phase at Degreed has been working out where human oversight comes in, says Burns. “AI can help you be productive, can analyse data more quickly than a human, but the results are not always accurate. So we’re teaching employees to use it as a starter, a foundation, but then there’s a quality control process you need to go through.”

There are no formal policies as yet, but as a guideline, employees should avoid putting proprietary data into generative AI as this carries the risk of data protection breaches. “It’s about asking ChatGPT for information rather than giving it,” she adds. “What it has been useful for is helping developers come up with taxonomies, mapping skills to roles, and personalise learning journeys on its own platform.”

Burns argues that the rate at which AI is becoming a mainstream part of the workplace means companies must act sooner rather than later. “It’s taught us that the concept of lifelong learning is real,” she says. “Before, people thought it was a nice- to-have but didn’t really understand the necessity of it. But with the tech advancing so quickly, if you’re not continuously learning, you will be left behind.”

L&D job opportunities on Personnel Today

Browse more L&D jobs