Animoca co-founder Yat Siu says one AI-era job skill will dominate

Yat Siu, co-founder of Animoca Brands, says AI is going to change what skills are most valuable in the workforce. That sounds simple, but it is the kind of shift that quietly rewires entire hiring pipelines, training budgets, and even job titles long before most teams catch up.

In Siu’s view, the AI era is not just adding a new tool to existing roles. It is re-ranking skills. The best way to understand what he is pointing at is to look at how work gets evaluated once AI starts doing more of the mechanical parts. When AI can draft, classify, summarize, search, and assist at scale, recruiters and managers usually stop rewarding people for raw output alone and start rewarding people for the judgment around output: what to build, how to validate it, how to fit it into a real process, and when it is wrong.

That matters because skills are not just personal development goals. They are a proxy for risk, velocity, and cost. If your company hires for “AI users” but what you really need is “people who can direct AI toward useful outcomes,” you end up with a mismatch. You might hire someone who can operate tools but cannot reliably translate business intent into working workflows, or you might staff roles with the right technical familiarity but the wrong understanding of how to measure quality. In practice, that creates friction across teams, because AI-assisted work often changes who owns decisions, who reviews results, and who signs off.

For boards and exec teams, the second-order effect is that AI-era skill shifts can stress existing org design. Job descriptions tend to lag technology. Teams write requirements around the last wave of change, then feel surprised when candidates do not match. The result is a cycle: you keep searching, time-to-hire stretches, and then you lower the bar or scramble for interim staffing. Siu is basically warning against that lag, even if the story itself sticks to the core claim: AI will change which skills the workforce values.

There is also a cultural layer. When AI changes what is rewarded, it changes what gets taught. Companies that rely heavily on “learning by doing” may find that opportunities for growth shrink for people who do not already have the new pattern recognition. Meanwhile, training programs that only cover tools, rather than decision-making, may become expensive and underwhelming. The point is not that tool literacy is useless. It is that tool literacy alone often stops being the differentiator once the market catches up and most candidates can get competent quickly.

You can also think about incentives. Hiring managers and HR teams are judged on throughput and retention. In an AI transition, the fastest way to hit numbers is to bring in people who can ramp on AI-enabled workflows quickly. But that can bias decisions toward whoever shows immediate productivity, not whoever has the strongest judgment and error detection skills. If executives do not adjust evaluation methods, they can accidentally overvalue the traits AI makes easy to imitate and undervalue the traits AI makes harder to automate, like framing problems, setting constraints, and auditing outcomes.

The regulatory backdrop, while not detailed in the CNBC item, is part of the real-world context execs have to factor in. Across many jurisdictions, policymakers have been paying closer attention to AI systems, data handling, transparency expectations, and accountability. Even without specific details in this source, the broader implication is straightforward: as AI becomes more embedded, the bar for defensible processes rises. That tends to elevate skills like documentation discipline, governance awareness, and the ability to align AI use with compliance requirements. Those are not “sexy” skills, but they can become make-or-break capabilities when scrutiny increases.

So the strategic stakes for leaders are immediate. Siu’s core claim is that AI will reorder the skill hierarchy in the workforce. If you run hiring, compensation, or capability planning, you cannot treat this as a training department problem alone. You have to treat it as an operating model problem. The companies that adapt will find it easier to build teams that can steer AI, validate outputs, and improve processes. The ones that ignore the re-ranking risk filling roles with the wrong skill mix, then paying for it later through churn, rework, and slower execution.