In certain instances, we may earn revenue when a reader makes a purchase or completes a form through our pages or on a partner’s website. We adhere to a strict 
There is a growing realization among workers that AI won’t replace humans — but those who harness AI will outperform those who don’t.
Take, for example, a recent Boston Consulting Group’s think tank experiment on 480 consultants, which showed up to 49% improvement on complex tasks outside their skillset when they could use AI as a tool.
Techopedia speaks with Vladimir Lukic, Managing Director and Senior Partner at BCG, to discuss how generative AI is entering the workforce.
We discuss how non-experts can perform complex tasks such as data science and how experts can level up with AI assistance.
We also cover the “exoskeleton effect,” how AI changes coding, and how managers can plan for an AI future.
Key Takeaways
- Generative AI boosts workplace productivity — but demands human oversight.
- Vladimir Lukic of BCG discusses the “exoskeleton effect” of AI helping both novices and experts.
- From coding to analyzing data sets, AI can reduce work from days to minutes.
- However, processes are needed to keep critical thinking alive in AI workflows.
- Leaders must use AI hands-on rather than just say: “Use it.”
Introducing the ‘Exoskeleton Effect’ of AI
A recent experiment conducted by BCG, along with scholars from Boston University and OpenAI’s Economic Impacts research team, led to what Lukic calls the ‘exoskeleton effect’ — AI as a tool that empowers workers to perform better and do more than a human or GenAI can on their own.
Lukic said:
“Think of GenAI as an exoskeleton. It can suddenly let a non-coder write code, but once you remove the tool, that person can’t do it alone. They haven’t learned to code.”
Research at BCG found that consultants with no coding background could perform advanced tasks using GenAI. But that proficiency (as you might expect) instantly vanished without the AI tool.
Despite novices showing dramatic improvements, experts benefited, too. Vlad points out that while GenAI levels the playing field for beginners, it also significantly accelerates and enhances an expert’s capabilities.
“Even our experienced data scientists got better at using these tools. People without backgrounds had enormous gains, but skilled folks also performed higher with AI augmentation.”
A big reduction in friction is GenAI’s natural-language interface allows novices to generate entire blocks of code.
However, the experiment revealed that verifying correctness relies on human judgment and partial knowledge of programming fundamentals.
“If you’ve never written code, you can prompt ChatGPT for a script, which hands you 30 lines of Python. But if you don’t know how to double-check it, you’re lost when it’s wrong.”
Why Outsourcing Everything to GenAI May Erode Critical Thinking Skills
Vlad highlights that fully outsourcing tasks to GenAI can undermine learning. Instead, organizations should structure processes so employees actively check and refine AI outputs, retaining critical thinking.
He added:
“We tell clients: don’t automate everything. Insert ‘purposeful toil.’
“Force some steps in the workflow where juniors must break the code into smaller parts, or they have to get a manager’s review. That’s how you build true intuition.”
Vlad shows GenAI’s strengths in quickly generating a sweeping overview of complex data. Yet, subtle insights and creative connections still rely on human expertise, illustrating how AI and human teams best succeed in tandem.
“We had times when the AI summarized 50 interviews and gave us a more comprehensive initial list of themes than any single analyst. But it still took a human to catch certain nuances it missed.”
Why Process Design Matters as Much as the Technology
The experiment suggests a massive expansion of tasks employees can tackle with AI support. Vlad advises leaders to rewrite job requirements and performance metrics, focusing on “AI synergy” rather than domain skills.
“It’s not about how many coders you hire. You need to consider who can fully leverage GenAI. Suddenly, that shapes recruiting, internal mobility, and how you form teams.”
While GenAI might slash execution time, organizations forfeit the efficiency gains if they cling to legacy checkpoints or bureaucratic sign-offs. Vlad urges executives to redesign workflows rather than just dropping AI into an outdated process.
We’ve seen processes cut from ten days to two minutes. The danger is if you keep the same approvals and committees, you’re back to a ten-day turnaround.
“That’s wasted potential.”
Vlad stresses that top leaders must adopt GenAI hands-on.
“Executives can’t just say: ‘We’ll do AI’, without trying it themselves. Whether for a quick draft of a board memo or analyzing data, they need that personal intuition for what the tool can and can’t do.”
At the same time, Vlad pointed out two risks:
“People love that GenAI boosts their confidence and speed, but it also raises questions: what if I start losing the core skill the AI’s handling? We need to plan for that.
“[At the same time] you can’t avoid AI. If you don’t provide a safe, responsible way to use GenAI internally, people will just go around you, using public tools and risking data exposure.”
The Bottom Line
Vlad’s observations suggest that GenAI can redefine how we think of “skill” in relation to the workplace. AI can offer dramatic performance boosts for both novices and experts, but it also reveals a critical gap between immediate capability gains and genuine skill-building.
Through structured processes, thoughtful leadership, and an eye on workflow redesign, organizations can unlock GenAI’s potential — while still nurturing the human superpowers that remain essential.
