Cybernetic teammates: how AI is (actually) reshaping work

Have you ever wished you could download and assimilate new knowledge instantly, like Trinity in The Matrix? Or maybe just absorb the expertise of the communications team while you write a technical report? Well, a new study suggests we might be closer to that reality than we thought. We recently came across one of the most revealing windows into how AI will impact our day-to-day work. It is a paper published by Harvard Business School researchers titled "The Cybernetic Teammate", and it details a carefully designed experiment with 776 professionals at Procter & Gamble. The findings are, frankly, hard to ignore. And it’s more exciting than you might fear.

How the experiment actually worked

Instead of trapping university students in a lab, the researchers took this experiment into the real world. They set up a virtual product development workshop, enlisting those 776 experienced professionals from both commercial and R&D backgrounds.

The participants were randomly split into four groups: individuals working alone without AI, two-person teams working without AI, individuals working with AI, and two-person teams working with AI. To ensure the results were practically useful, everyone tackled genuine product innovation challenges for their actual business units. They were not just testing a new tool in a vacuum; they were doing their real jobs, just with varying levels of human and digital support.

Crucially though, the researchers ran this experiment between May and July 2024. In the AI universe, that is practically ancient history. When you juxtapose the study's timeline against the massive collaborative AI developments that have rolled out since, such as Claude Cowork and OpenClaw, it is highly likely that these impressive findings are just the baseline. If AI was already this capable of being a Tier-1 teammate back in mid-2024, imagine what it can achieve with today's technology!

So, what happened when the results came in?

1. The solo worker becomes a dynamic duo

The headline finding is a showstopper. The researchers found that one person working with AI matched the output of a two-person team.

That is a literal doubling of effective productivity! It forces us to rethink the very nature of collaboration. AI is no longer just a fancy calculator or a glorified search engine; due to its access to vast amounts of knowledge and increasing ability to reason about solutions, it is stepping into the role of a collaborative partner capable of pulling its own weight. And remember, this was in 2024.

2. Smashing through professional silos

Usually, when an R&D professional tackles a problem, they lean heavily on technical solutions. When a commercial professional looks at the same issue, they focus on the market. It is human nature to stick to what we know.

But the study showed that AI-assisted professionals worked faster, produced higher quality work, and crucially, integrated perspectives outside their area of expertise much more effectively. In other words, the AI helped specialists think like generalists blending commercial and technical know-how into beautifully balanced, holistic solutions. This also helps to explain the one-person-team phenomenon more easily.

3. Levelling the playing field

If you are early in your career or simply tackling a task outside your usual wheelhouse, the learning curve can be notoriously steep. Not anymore.

The experiment revealed that AI assistance essentially closed the performance gap between novices and experts. It levelled the playing field, allowing less experienced professionals to perform at a standard previously reserved for industry veterans. Expertise is no longer bottlenecked by years of experience. Coding? Design? Mathematics? …Kung-Fu? It’s all up for grabs now.

4. The unexpected joy of being AI-enhanced

Here is the finding that really caught our attention. We often worry that new technology makes work more isolating, stressful, or robotic. And of course, that AI will leave us jobless.

Yet, people working with AI reported significantly higher levels of excitement, energy, and enthusiasm. Even better, they experienced lower levels of anxiety and frustration compared to those working without it. It turns out that having an endlessly patient, highly knowledgeable digital teammate actually makes your job more fun.

More productive and more fun. That's not the doom-and-gloom narrative we usually hear. And honestly, I have to agree.

As I’ve gotten slowly more comfortable with AI and its abilities over the last few years, I’ve found myself increasingly excited by what it makes me suddenly capable of. I can now create images that used to only live in my head (I’m not a great visual artist). Just the other day I coded an online tool, and I don’t know the first thing about coding! And if I want, I can take a day to create a visually stunning AI short film (actually, why not, I’m going to do that).

Part of me is still scared that AI will soon be able to do everything I can do behind a computer screen. But the more I use it, the more I think that fear might be asking the wrong question. The real frontier is not what AI can do instead of us. It's what we can now do that was previously unthinkable or simply out of reach. Things that were once too expensive, too technically complex, or too time-consuming for a single person to attempt. That's where human creativity, fuelled by AI as a force multiplier, becomes unbounded. And that kind of creativity might well fetch a premium.

A resounding online response

The broader response to the 2025-study has been overwhelmingly positive, with many experts viewing it as a watershed moment for AI in the workplace. The consensus is the research validates the shift from seeing AI as a mere automation tool to embracing it as a collaborative partner.

Conor Grennan, Chief AI Architect at NYU Stern School of Business, went so far as to call it "the MOST IMPORTANT study on AI in 2025" in a widely shared post, praising the stunning results. Similarly, industry commentators like Deborah Ko have noted that the findings prove "AI isn't the enemy of collaboration, it might be its next evolution." Ko highlights that AI can show up "not just as a search bar, but as a thinking partner" (Medium). Far from sparking fear about job replacement, the study has ignited excitement about how AI can make our daily work more dynamic and inclusive.

Evidence keeps mounting

Since the P&G experiment wrapped up, the AI landscape has evolved at breakneck speed, but newer research continues to back up these early findings. A January 2026 field experiment conducted by INSEAD researchers at a European technology firm took the premise a step further. Studying 316 employees, they found that those using generative AI did not just work better alone, they actually became significantly more central in their company's collaboration networks. The study showed that AI acts as a "translator" between different departments, dramatically increasing knowledge sharing between specialists and generalists. Far from isolating workers, the latest data confirms that AI can actively rewire and enhance human-to-human teamwork. The key is to embrace it as a tool, rather than shying away from change.

Ready to meet your new AI teammates?

The message from from these studies is clear: AI is not just a tool to save a few minutes here and there, it is a fundamental shift in how we work, learn, and collaborate.

As a researcher, you sit in a uniquely powerful position here. You already know how to evaluate evidence, think critically, and communicate complex ideas. Add AI to that skill set (used properly) and the impact on your research, your grant applications, your papers, and your outreach could be significant. The question isn't really whether to use AI. It's how to use it well, without cutting corners or compromising your integrity.

We can help you work better with AI in two ways

🧑‍🏫 AI for Researchers | live workshop

Want something tailored to your team or institution? Our AI for Researchers workshop is available year-round for groups. Get in touch to find out more.

🎓 AI for Researchers | online course

Our fully updated online course is currently closed for enrolments. The second intake will open in the second half of 2026. Join the waitlist to be the first to know when doors reopen.

The Cybernetic Teammate study showed that the performance gap between experienced and early-career researchers narrows significantly when AI is in the mix. But only if you know how to use it well.

That's what we're here for.