The Reward for Eating Pie is… Eating More Pie

Executive Summary:  The “More Pie” Paradox. In 2026, AI is finally delivering measurable productivity gains, but at a hidden cost. Research shows that without a clear strategy, AI doesn’t just save time—it expands workloads, blurs professional boundaries, and increases error rates through forced multitasking. For leaders, the goal shouldn’t be “more pie,” but smarter portions.

I was recently preparing for my upcoming TechShow when a colleague shared something a university lecturer once told him. We were discussing how to help teams understand the rewards and benefits of integrating AI into their workflows. His response? “The reward for eating pie is eating more pie.”

I couldn’t have put it better myself.

We are currently trapped in a continuous loop. Everyone wants to introduce AI to save time, improve productivity, and cut costs. This is particularly relevant right now, as we navigate a “hidden” economy where unemployment continues to rise (specifically in the UK) and companies remain reticent to hire new talent. If hiring has stopped, but the work hasn’t, who is doing the job?

The answer is: probably you, with the help of AI.

The Paradox of AI Productivity

This is the core conclusion of ongoing research by Aruna Ranganathan and Xingqi Maggie Ye at Harvard. They found that when it comes to AI, the reward for eating pie is, indeed, more pie.

In an eight-month study of generative AI adoption at a U.S. tech company, they found that employees worked at a faster pace, took on a broader scope of tasks, and actually extended their working hours, often without being asked. From a corporate perspective, this sounds great. More work, faster, for the same cost. But the researchers highlighted several “blurred lines” that we cannot afford to ignore:

• Policy Blind Spots: An engineer might be hyper-aware of regulations and safety protocols, but “Sally from Sales,” who is building her own automations to make her life easier, might not be.
• Unsustainable Workloads: While people enjoy experimenting with AI initially, it is leading to a volume of work that isn’t sustainable. This risks long-term damage to morale and team culture, especially when some team members embrace the tech while others reject it.
• The Multitasking Trap: AI encourages multitasking. While this sounds efficient in principle, the reality is that as multitasking increases, the rate of errors increases exponentially.

 

Adoption Without Strategy

Why am I telling you this? Because a business’s adaptability to AI shouldn’t just be about how to adapt, but also what we do once the transition happens. What we see in this research is a clear example of adoption without strategy.

We are finally starting to see the first real signs of AI-driven productivity growth in early 2026. It feels strange that it took nearly four years since the launch of ChatGPT to see these gains, but they are here. Research from the Federal Reserve of St. Louis, “The State of Generative AI Adoption in 2025,” found that industries saving the most time with AI were indeed seeing unusually fast labour productivity growth.

But we have to ask: at what cost? If this productivity is built on blurred roles, unsustainable workloads, and high error rates, is it a real gain?

The Accenture Approach: A Warning Sign

While some are struggling with strategy, others are leaning into force. Just this week, news broke that Accenture has begun monitoring staff use of AI tools as a metric for top-level promotions. They are essentially pushing reluctant employees to adopt the technology by making it a performance requirement.

In my opinion, this is definitely not the way to go.

If you want your employees to truly embrace AI, you are better off listening to their concerns. You might actually learn something in the process. As I mention in my framework, the number one skill a leader needs in the AI era is the capacity to listen. By forcing your teams to incorporate AI, you aren’t just ignoring their feedback, you are simply pushing them to work faster, and likely, to make more mistakes.