Understanding the Impacts

While many have long lamented the potential for AI to “take away their jobs,” the reality is far less bleak. Their jobs are likely to still exist, but they’ll look far different and require different skills than in the past. That’s as true for information professionals as it is for a wide range of other professions.

“AI isn’t replacing information professionals entirely—it’s reshaping what they do and how they do it,” says Lacey Kaelani, CEO of Metaintro, a job search engine that runs on open source data processing. “The valuable skills now are architecture, fact-checking, and teaching people how to ask smarter questions,” she notes. “The old tasks of cataloging and organizing, which were major responsibilities in the past, are becoming background work now done by AI tools.” 

The most important skill in the labor market will be AI literacy, including prompt engineering, according to Kaelani. “Not programming, but knowing when to push back on the information the model gives you and how to defend your reasoning to others.” Ethical filtering will also be important, she says: “Catching the bias, errors, and blind spots that AI consistently misses.”

Many have had to learn on-the-fly, as neither educational institutions nor organizations had the knowledge, experience, or expertise to understand the skills needed, let alone to train employees in them. Shifts are being made, however.

Thomson Reuters’ “The Future of Professionals” report makes an important point: The biggest barrier that organizations are facing today with the advent of gen AI and a related plethora of new tools and technology isn’t the technology itself—it’s helping people learn how to use it. The numbers are striking:

• 71% say they don’t feel prepared to use AI at work.
• 46% say there are skill gaps on their teams.

The news isn’t all grim, though. The report also says that those who have strong AI knowledge are 2.8 times more likely to deliver business value. What that means for organizations and individuals is the need for reskilling and upskilling to effectively adopt these tools and adapt to a new way of working.

What to Learn

Hanna Parkhots is a data collection project manager at Unidata. She has been in this role for nearly 3 years, overseeing data collection projects conducted both in-house and through crowdsourcing platforms. Upskilling and reskilling need to focus not only on how to use AI, but on critical AI literacy, she says. “It extends beyond typing a prompt into a gen AI software. It’s recognizing the mechanics behind such technology, where it falls short, but fundamentally, where it stands with respect to ethics.” 

Parkhots says her teams have a specific focus on “checking AI-created data with human-in-the-loop verification because our internal audits reveal that as much as 15% of AI-created answers in some datasets have faint errors or biases.” The ability to do this effectively, she notes, “requires a keen eye for detail, along with a solid grounding in classical research methodologies.” In addition, “professionals also have to be information systems architects, where they craft prompts and workflows to elicit maximum quality output from AI with a concomitant reduction of the harms of misinformation.”

Brown breaks down the required skills to succeed in this AI-driven era into three key categories: 

• AI literacy—how it works
• Workforce integration—how to apply it to your work
• Change navigation—how to scale safely

Employees need to understand the “strengths, limits, and risks of gen AI tools,” Brown says, and understand “when human oversight is non-negotiable”; for instance, for guarding against bias, hallucinations, and context gaps. They have to understand how to use structured prompting frameworks and how to embed AI into their daily tasks, “like research summaries, metadata tagging, knowledgebase updates, or user-facing guides.” Importantly, AI needs to be paired with “existing library systems, citation management, and databases rather than treating it as separate,” Brown cautions. 

Employers, Brown continues, have to be able to assess tools and trends quickly, set policies for responsible use, and lead both “peers and patrons through adoption by modeling best practices.”

Clearly, these are no simple mandates. Employers are seeking ways and sources for educating both themselves and their employees—and staying up-to-date with ongoing changes.

Where to Learn

One of the challenges for both organizations and individuals today is that this technology is so new that schools at both the K–12 and higher education levels are having a hard time keeping up. This means that even recent graduates may not have the timely, relevant, and applicable information and training they need to hit the ground running. That’s where reskilling and upskilling can come in, and, fortunately, there are places to turn. 

There’s a lot that individuals can and should be doing to build these skills. Jason Hishmeh, co-founder of Varyence and Increased with expertise in software development, cybersecurity, cloud infrastructure, and AI, recommends that people “start with platforms like Coursera, DeepLearning.AI, and OpenAI’s documentation, which don’t charge a lot and are often great jumping-off points.” But, he adds, technical training isn’t enough. In addition, “teams need to understand how the tools work to find better answers for users, what biases they might introduce, and how to ensure their processes are compliant while using them.”

In this rapidly changing environment, it’s important to recognize the value of teaching yourself, says Kaelani. “The people getting ahead are teaching themselves through free online courses and experimenting with tools,” she explains. “Academic programs are lagging, but professionals who treat AI like learning a new language—and practice with it every day—are going to be the ones who see salary raises in their future.”

But organizations, of course, also have a role to play—a role that’s necessary, according to Hishmeh. “Providing people with sandbox environments, micro-credential support, and activities like hackathons is important so that they can test AI tools without fear,” he says. “The best upskilling happens in real-time, tied to real challenges.”

Brown agrees: “Self-study is valuable, but it should be paired with peer review or community practice, because working alone with AI increases risk of human errors.”

“While individual effort is crucial, employer support is a game-changer,” says Parkhots. Her company has allocated a dedicated budget for professional development, allowing her team to enroll in courses and attend workshops. “We’ve seen a measurable return on this investment, with project efficiency increasing by an estimated 20% after my team completed a specialized course on prompt engineering.”

This type of support, continues Parkhots, is not just financial; it’s about providing time and resources for employees to train without impacting their core responsibilities.” This, she suggests, “could take the form of dedicated ‘innovation hours’ each week or a fund for external certifications.” 

Parkhots notes that educational institutions are gradually catching up. “We’re also witnessing an increasing number of specialist courses being created, not just within broad-based computer science programs, but even within library and information science programs,” she says. She points to courses like Information Ethics in the Age of AI or Prompt Engineering for Knowledge Management. These types of courses, she says, “will fill the gap between legacy information skills and the requirements of the AI-led world.”

In the meantime, though, Brown says, many “will find more immediate value in external programs and professional associations that are actively updating AI modules.”

It’s an environment of both uncertainty and excitement. Nobody has all of the answers—and certainly not the “right” answers. Success lies in open-mindedness, comfort with ambiguity, and the willingness to explore and experiment.