How AI Makes Expertise More Valuable

↓

​Learn more about the People Analytics course →​

Friends,

Before we get going... let me do a quick plug!

Every AI conversation seems to eventually boil down to the same question:

Which jobs survive, and which don't?

This weekend, I even met an AI engineer whose job is, quite literally, to help automate parts of her own role.

Can you believe it?!

But let's be clear, jobs will not JUST disappear out of the blue. I think the more interesting question is how the work will be augmented by AI.

That's exactly what we're digging into on September 23 at the Toronto People Analytics Group.

​Jin Yan from Revelio Labs, our event sponsor, will break down what's actually happening at the activity level:

• Which tasks AI is already reshaping
• Which roles may be most exposed
• Where human skills are becoming more valuable
• And why job titles alone may be the wrong way to think about workforce strategy

And Revelio has a HUUUUUGE dataset to back it all up.

This is a live event in Toronto, and space is limited.

​See the event details and reserve your spot →​

Now, back to the newsletter! And today, we will talk about Revelio data itself!

Every pitch deck and panel discussion about AI seems to repeat the exact same narrative:

• AI writes
• AI codes
• AI analyzes
• AI research
• AI builds slide decks

Because AI allows a single person to self-extend so much every single day, the common assumption is that employers will shift toward hiring generalists.

Indeed: Why hire three specialists when one AI-augmented generalist can cover the entire workflow?

It sounds logical. But when you look at actual labor market data, a very different pattern emerges.

Recent workforce research from Revelio Labs analyzing approximately 75 million technology job postings suggests that tech employers aren't hunting for broad "jack-of-all-trades" skill profiles.

Instead, they appear to be demanding much deeper experience, not broader.

Say NO to Skill Laundry List

For years, technical job descriptions read like wish lists.

Managers and HR would add 30 different required technologies onto a single posting, hoping to land a unicorn candidate who could handle everything from database administration to frontend styling.

That era appears to be cooling off.

According to Revelio Labs, since late 2024, the average number of skills required in tech job postings fell from roughly 30 down to 21.

That is a 25% drop in skill breadth.

At the exact same time, the average required experience associated with listed skills rose from 4.6 to 4.8 years.

Employers are trimming the superficial requirements and doubling down on specific experience.

From General to Specific

This structural shift becomes clearer when you look at which skills are disappearing from job posts.

In 2023, general software development appeared in 25.4% of tech job postings. This year, that number dropped to 16.5%. Application development took an even steeper plunge, falling from 14% to 6.4%.

Meanwhile, highly specific capabilities like Java API development, Python development, Cloud deployment, Machine Learning, AWS Data Pipeline, and advanced networking have steadily gained relative weight.

There is a big difference between saying "I can write code" and saying "I have six years of experience optimizing this specific data pipeline architecture."

AI tools make basic general execution easy to fake:

Anyone can generate a functional Python script or a basic React component in thirty seconds. Because basic execution is becoming cheap and accessible, saying "I can develop software" no longer differentiates a candidate. Deep expertise solving specific, complex technical problems carries the real premium.

Knowing People Analytics means moving beyond simply reporting which skills are increasing or decreasing in job postings. It requires understanding what those structural shifts mean for actual workforce strategies, compensation models, and talent pipelines.

That is why I always say that People Analytics is not a function.

It is a tool.

​→ Learn more about the People Analytics course​

Judgment Is Expensive

The shift toward deep experience shows up most aggressively in system architecture and data capabilities.

Revelio Labs found that required experience for data organization and research skills jumped from roughly 2 years in 2023 to 5 years today—a 150% increase. Requirements for enterprise architecture design rose from 5 years to 6.5 years.

Why would AI increase the premium on years of experience?

Because there is a vast difference between getting an answer from an automated tool and knowing whether it's the right answer.

AI can execute tasks faster than any human. But execution is useless without judgment.

Someone still needs to:

• Understand how the overall architecture hangs together
• Instantly spot subtle errors or error logic in automated outputs
• Understand complex upstream and downstream dependencies
• Make trade-offs between performance, security, and cost
• Frame the initial problem correctly before prompting a tool
• Recognize when a process should be automated or not

When execution gets easy, judgment becomes the bottleneck.

And judgment is almost entirely a function of experience and human intelligence.

The AI Hiring Signal

When Revelio Labs looked specifically at job descriptions that explicitly referenced AI usage in their responsibilities, controlling for role, seniority, industry, and overall description length, they found the exact same structural signature.

These AI-inclusive postings listed approximately:

• 3% fewer required skills
• 3% more required experience

The difference here is modest, and we shouldn't overstate it. This data doesn't prove that AI adoption caused the entire labor market to specialize overnight. But the correlation is consistent across millions of records: roles adapting to AI tend to lean away from broad skill checklists and toward focused depth.

Using workforce data to decode changing skill requirements and translate them into concrete workforce decisions is precisely what we work through with enterprise clients.

​→ Explore our consulting services​

Asking the Right People Analytics Questions

Most organizations are evaluating AI through an overly simplistic lens:

"Which jobs will AI eliminate?"

That is usually the wrong question.

A more practical framework for People Analytics teams looks like this:

• Which skills inside our business are easy to augment?
• Which roles require rare, non-negotiable skill depth?
• Where is broad, surface-level execution losing market value?
• Where is judgment becoming a competitive advantage?
• How must our job architectures evolve to reflect the new work?

You can't only count skill occurances.

You need to evaluate the skill density holistically.

The Entry-Level Pipeline Problem

Here is the uncomfortable issue hidden inside this data.

If organizations demand fewer total skills but far more years of experience, what happens to early-career talent?

You cannot hire a senior engineer with six years of deep architectural experience unless someone, somewhere, gave them their first year.

If companies eliminate junior execution tasks through automation while simultaneously demanding experienced specialists, they break their own talent pipeline.

It's a structural workforce design problem.

If your organization automates away the work that junior employees traditionally used to build basic competencies, you must deliberately design new ways to build senior capability:

• Rebuilding formal apprenticeship structures
• Designing cross-functional internal mobility tracks
• Mapping explicit skill-progression pathways
• Creating structured environments where junior staff practice high-level judgment, rather than just executing routine tasks

The CHRO & People Analytics Playbook

If you are leading People Analytics or Workforce Planning, here are four adjustments to make right now:

• Stop treating every skill as equal: Differentiate clearly between broad operational capabilities (which AI easily augments) and deep specialized capabilities (which require years of domain context)
• Measure experience alongside skill supply: Knowing that 500 employees possess a specific skill tag in your HRIS tells you very little if the business actually requires 5+ years of production experience in that discipline
• Map how AI changes the skill bundle: Don't just track headcount losses. Track how tasks within a role shift, and which higher-order skills become critical once basic execution is offloaded
• Protect your experience pipeline: If routine junior tasks vanish, create explicit internal development pathways to teach judgment. Otherwise, your future senior talent supply will simply dry up

If you're trying to figure out how to analyze these skill shifts or restructure your job architecture for an AI-enabled workforce, that's a conversation we can have.

​→ Learn more about our consulting work​

Yes, there are more plugs now, but I truly hope you are enjoying the content.

Till next time!

K

Whenever you’re ready, there are 2 ways I can help you:

#1

If you’re still looking to get started in People Analytics, I recommend starting with my affordable course:

​Practical People Analytics: Build data-driven HR programs to 10x your professional effectiveness, business impact, and career. This comprehensive course will teach you everything from building an HR dashboard for business results to driving growth through more advanced analytics (i.e., regression). Join your peers today! ​
​

#2

If you are looking for support in your human capital programs, such as engagement, retention, and compensation & benefits, and want to take a more data-driven approach, contact me at Tskhay & Associates for consulting services. Or simply reply to this email!