When AI Becomes A Teammate

Over the past year, I’ve had hundreds of conversations with marketers — more than 350 through my lecturing work at RMIT alone. A common theme keeps surfacing: “ChatGPT has completely changed how I work.”

They tell me it’s compressing tasks that used to take days into hours. Hours into minutes. Often, these comments come with a tone of quiet confession — like they’re revealing a secret. And that’s because they often are. Most of the time, they’re not sharing this with their manager.

This is the reality of how generative AI is already reshaping knowledge work — quietly, informally, and profoundly.

Until recently, the evidence for this was largely anecdotal. Canva’s *2025 State of Marketing & AI report added some weight to the story: surveying 2,400 global marketing and creative leaders, it found that 85% are reclaiming the equivalent of a full workday every two weeks, and 20% are getting back 10+ hours per week.

That’s encouraging — but surveyed data always comes with a caveat. Most of us have been on the other end of those forms. It’s easy to tick a box. It’s much harder to observe real behaviour, in real teams, under real conditions.

That’s why empirical evidence — gathered through observation and experimentation — matters. And that’s what makes a ^new study published this week by Harvard Business Review such a milestone moment in the evolution of AI at work.

What Happens When AI Joins the Team?

The study looked at how AI performs not just as a tool, but as a teammate. Conducted inside Procter & Gamble, it involved 776 professionals — experienced R&D specialists, and commercial leads — working in teams on real-world business challenges across categories like oral care and baby products.

Some teams worked as normal. Others got access to GPT-4 or GPT-4o, with training to use it effectively.

Here’s what happened:

• AI made individuals as effective as teams. A solo professional with GPT-4 produced outcomes on par with a traditional two-person team (diagram 1)
• AI-augmented teams delivered better results. They were significantly more likely to produce top-quality solutions than those without AI (diagram 2)
• AI sped up work. Tasks were completed 12–16% faster — with more detailed, thoughtful outputs.
• AI made work feel better. Participants reported more positive emotions and less stress when working with AI.

So What?

This isn’t just about efficiency. It’s a shift in how work gets done — in how we collaborate, problem-solve, and tap into new kinds of capability.

AI isn’t replacing teams. But it is reshaping them. It’s moving from tool to teammate. Quietly at first, behind closed doors. But now, increasingly, out in the open — and backed by real evidence.

We’re only just beginning to understand what that means. But one thing is already clear: the future of work won’t just be human. It’ll be human + AI — working better, together.

‍