As organizations race to adopt AI, the question is no longer if but how employees will learn to use it. On the World Economic Forum’s Radio Davos, Workera CEO Kian Katanforoosh made the case that mentorship — specifically AI mentorship — will be what separates organizations that dabble in AI from those that build lasting capability.
The distinction is more urgent than it seems. A 2025 Cisco study found that 97% of CEOs plan to integrate AI, yet just 1.7% feel fully prepared to lead the transition. In short: ambition is high, but readiness is not. Further research on intelligent tutoring systems shows that AI-driven mentors can deliver measurable learning gains compared to traditional training.
Below are four critical takeaways from Kian’s conversation to guide workforces through adoption, scale learning velocity, and position their companies ahead of the competition.
1. The half-life of skills is shrinking… and that changes everything
Because the useful life of skills is shrinking, organizations can no longer rely on static training programs to keep their workforces competitive.
Kian stressed, “The pace at which people need to learn skills is higher than it’s ever been in history. A metric that shows how long a skill might be useful in someone’s career decreased drastically — from over a decade 40 years ago, to now just 2.5 years in digital areas.”
Kian referenced the World Economic Forum’s “skill half-life” metric to explain why one-off training programs won’t scale. He added, “If you have a skill, it will be useful just for two and a half years, and then you’re going to need something else.”
People should “learn how to learn” and “learn how to talk to machines.” These are not abstract slogans; they describe the day-to-day fluency organizations must cultivate if they want employees to remain innovative, productive, and competitive.
If technical and AI skills expire faster, learning and development strategies have to adapt to meet the pace and scale of skills’ turnover rate. Strategies must shift from episodic coursework to continuous, personalized learning pathways that are integrated into work itself.
2. AI mentorship multiplies learning velocity
A core thread of the interview was the promise of AI mentors: systems that assess someone’s current capabilities, set ambitious but realistic goals, and recommend the exact learning steps and projects that will move them forward.
Kian described the mentor role simply: “The first thing an AI mentor has to be extremely good at is assessing skills. Humans aren’t great at assessing their own skills or other people’s skills. Second, a good mentor needs to understand goals — and the best ones push you to dream bigger. And third, it’s about connecting the dots. Humans can’t keep up with which projects, courses, or tools are best right now. AI can.”
Kian’s perspective aligns with broader research on adaptive tutoring. Studies of intelligent tutoring systems and AI-driven learning assistants show measurable gains for learners exposed to personalized, adaptive guidance. AI mentors can scale one-to-one coaching and drive faster, more targeted skill growth than generic courses.
Organizations that invest in mentoring infrastructures — both human and AI — can speed skill acquisition and reduce the mismatch between business needs and employee capabilities.
3. “Learn by doing”: Embed AI into real work, not just training catalogs
Kian emphasized a practical point leaders often miss: employees learn fastest when AI is part of the day’s actual tasks.
“Enterprises used to teach with one-size-fits-all classes,” said Kian. “You’d send the same course to tens of thousands of people, and very few completed it. Instead, today, you put an agent on top of the latest courses, let it assess your strengths and gaps, and then it tells you: these are the three or five videos you need to watch this week. It’s 15 minutes, and you’re good to go.”
Rather than forcing everyone into the same course, an agent can assess a person’s strengths and gaps and suggest a few short, targeted activities that are directly relevant to the next project. “You don’t need to think where you should spend your time,” he said — the agent will surface a focused set of materials tailored to immediate work goals.
Employees who learn by applying AI in context build confidence and retention far more quickly than those in passive, classroom-style programs. Embedding AI prompts and automated guidance into workflows turns training into productive time rather than “time away” from the job.
If learning programs don’t connect to the work people do, the result will be low completion, low transfer, and little business impact.
4. AI can reduce human bias, but only when we design for fairness and oversight
Kian was candid about the tradeoffs. He argued the current resume-filtering funnel is blunt and often unfair, and suggested rethinking hiring pipelines when AI enables broader, validated assessment of skills.
“We’ve designed recruiting as a filter because it takes so much time to interview people,” Kian said. “But resumes are a bad asset for skill measurement in the first place. In an age where AI can interview everyone, the funnel doesn’t make sense anymore.”
AI won’t replace humans immediately or fully. Instead, the best approach is hybrid: let AI surface evidence and have people add human judgment where nuance matters. As he put it, “You figure out where AI is better than humans and where humans are still needed.”
This balanced stance reflects a broader debate: some surveys show younger workers already turn to AI tools for career advice — nearly half of some Gen Z samples prefer AI for career guidance — while analysts warn chatbots can miss nuance, foster overconfidence, or fail to challenge assumptions. Businesses must pair automated assessment with explainable models, clear feedback loops, and human supervision to preserve fairness and trust.
Still, Kian believes future research will align with this shift to AI in hiring, adding, “I truly think that as soon as a year from now, you’ll start seeing validation studies showing that AI is a better interviewer than humans.”
If an enterprise adopts AI in hiring, development, or promotion decisions, they must ensure explainability, bias testing, and human review are core requirements — not optional extras.
Building AI readiness now
Kian’s conversation on Radio Davos is a practical reminder: people, not just code, determine whether AI delivers business value. Build learning systems that adapt, mentorship that scales, and governance that protects fairness. As Kian urged, focus on “learning how to learn” and make adoption part of everyday work.
If you’d like to see an AI mentor in action, learn more about Sage, Workera’s AI mentor that combines skills verification, adaptive learning pathways, and contextual coaching to help employees move from awareness to fluency.