Q&A: How AI is reshaping the technology workforce

There has been an increase in jobs requiring an AI skill and, at the same time, feelings among a sizable section of the workforce that they feel inadequate when it comes to acquiring or demonstrating such skills. Another development is with how AI is reshaping the tech workforce.

To gain insights into this, Digital Journal spoke with Paul Farnsworth, President of Dice. Farnsworth discusses what hiring data reveals about the demand for AI skills, how AI is changing the work of engineering and IT teams, and why systems expertise and institutional knowledge are becoming more valuable as companies deploy AI across the business.

Digital Journal: AI has fuelled widespread concern about job displacement. What are employers’ hiring patterns telling us about how AI is actually reshaping the U.S. tech workforce?

Paul Farnsworth: We’ve spent the last year asking the wrong question. Instead of asking whether AI is replacing jobs, we should be asking how it’s changing the work people do.

That’s what hiring data is telling us. Our recent Tech Hiring Myths report found demand for AI skills has increased 380% since early 2024, AI and machine learning job postings have grown 173% year over year, and nearly three-quarters of tech job postings now require at least one AI-related skill.

That doesn’t tell us companies are hiring fewer technologists. It shows that these organizations are looking for people who understand how AI fits into the way modern businesses operate.

What’s interesting is that AI is showing up in three very different ways: 

• AI in products: Companies are building AI into new products or integrating generative AI into existing ones. 
• AI at work: Technology teams are enabling AI across the business, putting the guardrails in place so employees can use it securely and effectively. 
• AI agents: Organizations are beginning to deploy autonomous agents that can work across enterprise systems

Each of those shifts creates new work for engineers, platform teams and technology leaders. The work may change, but it never disappears. 

DJ: How is AI changing what companies need from engineering and product teams?

Farnsworth: We’re seeing two paths emerge. Some organizations are designing products around AI from the beginning. Others are taking products they already have and adding generative AI capabilities to them. Both approaches can be successful, but neither is really about the model itself.

The companies creating the most value are the ones that understand their own data and know how to connect it to AI in a meaningful way. That’s where engineers become incredibly important. Someone has to build the pipelines, prepare the data, create the APIs and make sure everything works reliably once it reaches customers.

Access to powerful AI models is becoming much easier. The harder part is knowing how to connect them to a company’s own data and use them to solve real customer problems. That’s why I see demand continuing to shift toward engineers who understand systems and business problems they’re solving over just having familiarity with the latest AI technology.

DJ: Beyond customer-facing products, how are companies using AI inside the business?

Farnsworth: Some of the most important AI work isn’t customer-facing at all. Technology teams are figuring out how to make AI useful for employees, whether that’s helping developers write code, supporting analysts’ work with data or making everyday business tasks more efficient. But none of that happens by accident.

Someone has to connect those tools to existing systems, decide what company information they can access and put the right guardrails in place. Otherwise you end up with dozens of disconnected AI tools, inconsistent practices and unnecessary security risks.

I think that’s one of the biggest workforce stories people miss. Internal technology teams aren’t just deploying AI anymore, as they’re a huge part in helping the rest of the business understand how to use it responsibly and effectively.

DJ: AI agents are getting a lot of attention. How do they change hiring needs?

Farnsworth: Unlike AI tools that primarily help employees work more efficiently, AI agents can take action within the workplace. An agent might pull information from one system, update another and trigger a workflow without someone manually stepping through each task. That’s powerful, but it also means companies have to think much more carefully about how those systems are managed and governed.

The people who will be valuable here aren’t necessarily brand-new “AI specialists.” They’re often experienced platform engineers or infrastructure teams who already understand how a company’s systems fit together.

That doesn’t necessarily mean companies need to hire entirely new teams. Instead, many organizations will focus on upskilling existing technical talent, particularly engineers and infrastructure teams who already understand their enterprise systems. As AI becomes more embedded across the business, employers will prioritize professionals who can integrate, manage and govern these technologies rather than simply use them.

As AI becomes more autonomous, institutional knowledge becomes more valuable, not less. If you’re going to let software act across your business, you want people who understand the systems it’s interacting with.

DJ: Are concerns about AI replacing entry-level tech workers being overstated?

Farnsworth: The concern is understandable, but I’d describe it differently. AI is going to automate some of the work that’s traditionally helped early-career professionals build experience. That means the path into a career may look different than it did five years ago.

But companies still need people who can think critically, solve problems and understand how technology fits into a larger system.

For entry-level professionals, simply knowing how to use AI won’t be enough. Using AI is quickly becoming table stakes. The differentiator is judgment. Can you tell when AI got something wrong? Can you explain why you changed its output? Can you understand how your work fits into a much larger product or platform? Those are the skills employers will be looking for as they hire the next generation of technical talent.

Those are the skills that are becoming more valuable, not less.

DJ: As AI becomes part of everyday business, how should companies think differently about hiring?

Farnsworth: The mistake I see companies make is treating “AI talent” as a single hiring category.

The skills you need to build an AI-powered product are different from the skills required to roll AI out across your workforce, and those are different again from the expertise needed to manage autonomous AI agents. 

The starting point shouldn’t be, “We need AI people.” It should be, “What are we trying to accomplish, and what capabilities do we need to get there?”

At the same time, companies shouldn’t overlook the people they already have. Employees who understand the organization’s systems, customers and data often have a huge advantage because they already know how the business works. Teaching those teams how to apply AI is often more valuable than hiring someone who knows AI but doesn’t understand the company.

The biggest shift I see is that the premium is moving away from model expertise and toward systems expertise. As AI becomes more accessible, the people who understand data, infrastructure and business context are going to be the ones creating the most value.