Organizations must rethink skills to realize AI ROI

The adoption of AI tools in the workplace shows no sign of slowing down. With 88% of businesses using AI in some capacity, companies are now highly focused on trying to measure the ROI from their investments in the technology.

Companies deploying AI tool stacks across their teams aim to boost productivity by automating routine tasks, synthesizing data and improving efficiency, reducing both costs and time-intensive labor.

But there is friction. Half of London businesses say their workforce does not currently have the skills needed to meet their organizations AI requirements. Companies are now facing a costly problem, they need a return on the investment in AI and it’s going to take a revamp of skills and job roles to do so.

Businesses need to take a head-on approach to solve this challenge by balancing technical knowledge and creativity as well as business understanding.

It is not just technical IT skills required for back-end ops, but also digital literacy, critical evaluation of AI outputs and workflow integration. These are becoming essential at every level, from entry level to C-Suite level.

Job Descriptions need updating

As the digital skills gap widens, it is unsurprising that, in the UK, one in four workers worry AI could replace them. In reality, the labor market is splitting into two distinct tiers.

The professional tier is centered on blended roles that rely on human judgement and recognizes that AI tools are only as effective as the people guiding them. By contrast, the democratized tier sees repetitive tasks handed over to AI tools entirely.

With most roles sitting in the professionalized tier, maximizing AI efficiency requires critical thinking and judgement that the technology itself cannot supply. However, if those expectations are not written into the role, employers end up overestimating the capabilities of their employees.

Nowadays, many roles have become blended with technology, and, except for technical roles, many necessary AI capabilities are often summarized into a single generic bullet point on technical proficiency. What we are seeing is a reshaping of what employers value. This reshaping is driving the emergence of creative technologists.

A creative technologist is someone who understands technology, creativity and strategy, and organizations that actively empower this role are more likely to pull off digital transformation successfully.

Today, many teams are full of creative technologists who don't realize that's what they've become. The forces driving this are generative AI becoming a daily tool, rising content demands, and increased cross-discipline collaboration.

But employers need to ensure job descriptions match this reality. A blended job description needs to name the AI tools the person will actually use, then spell out the skills needed to use them, moving things like critical thinking from generic bullets into specific expectations tied to those tools.

If employers get that right, they will be hiring people who can capture the productivity gains and not just operate the tools. AI success requires people who combine technical expertise, industry know-how and creativity.

Without more blended roles, businesses are going to struggle to see the return on investment and ensure they are using the right tools for the right job.

Encouraging trialing, testing and tryouts

Effective learning can only happen in an encouraging environment. Getting comfortable with prompting frameworks makes AI tools feel less daunting, and that comfort is often what surfaces where AI can actually help within the organization. Experimentation matters even when a given approach doesn't get pursued further. Sometimes the most useful lesson is learning what not to do.

One of the biggest barriers to that experimentation paying off is dark data. Most organizations sit on enormous volumes of information, but a large share of it lives in documents, emails, PDFs, and other unstructured formats that are hard for AI systems to interpret.

That means even well-designed models often work from an incomplete, inconsistent view of the business which leads to unreliable outputs and initiatives that stall before they scale.

Moving past that requires turning dark data into something usable that can build upon the structure, context, and governance AI systems need to act on the data with confidence. Once that foundation is in place, AI becomes far more reliable, and far easier to move from isolated pilots into real production environments.

The human edge

AI can streamline planning, but people still drive execution. As tools take on more routine workflow, organization and documentation tasks, the skills that remain most valuable are the ones AI cannot replicate such as judgement, storytelling, creative thinking and the ability to lead people through change.

Demand for AI-fluent specialists has created a new kind of empowered generalist who can read AI trends as fluently as they manage stakeholder relationships and understand customer pain points.

AI may automate workflows and documentation, but it cannot lead a team, interpret context with human nuance or fully understand a company’s mission.

Where wearing many hats within one role was once seen as a drawback, AI is making it one of the most efficient ways to work. It gives employees access to skills that would previously have taken years, and several role changes, to acquire.

Rather than simply eliminating roles, AI is also creating new ones centered on managing the human-AI relationship, including AI evaluators and human-in-the-loop reviewers.

In this sense, AI is not replacing jobs so much as reshaping them, with the potential to reduce workforce inequality by democratizing access to new skills.

The skills gap is a hurdle many organizations are now facing and both employees and employers must experiment to adapt.

We've featured the best AI chatbot for business.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit