.svg)
When generative AI first became widely accessible, there was a clear expectation and direction from Leadership mandating it be thrown everywhere possible - productivity would increase. Tasks would take less time, workloads would reduce, teams would become more efficient. But for most organisations, that’s not the reality on the ground. Instead, something else emerged.
The generation problem
GenAI gave people the ability to create more, faster. More content, more ideas, more documents, more options. But it didn’t remove the underlying work. If anything, it increased it. Because generating something is only the first step. That output still needs to be reviewed, refined, validated and integrated into real workflows. So instead of reducing effort, GenAI often just shifted it. It gave everyone a generation superpower, but not a completion superpower.
Why agents change the equation
The real shift is now happening with agents. For the first time, we’re seeing AI move beyond generating outputs and into taking actions, completing tasks and executing workflows. This is a fundamentally different capability. An agent doesn’t just suggest what to do next - it does it. That’s where real productivity gains start to emerge. Not from producing more, but from removing steps entirely.
The new risk: Solution overload
At the same time, a new challenge is developing. Because building with AI has become easier, everyone is building something - internal tools, automations, prototypes….quick fixes. We’ve seen this first-hand. You can spin up something that’s 80% complete in under an hour.
But that last 20% - the part that actually matters - takes significantly longer. Making it secure, making it scalable, integrating it into existing systems and ensuring it works reliably. You can build a working prototype in under an hour. But it can take 6-8 weeks to get it production-ready.
That gap is where most organisations are underestimating the challenge.
Trend vs shift
Right now, the trend is everyone building their own tools, decentralised experimentation and individual productivity hacks. But the long-term shift will look different. Organisations will move towards centralised systems, dedicated teams building scalable solutions and shared infrastructure that the whole business relies on. Because while individual experimentation is valuable, it doesn’t scale.
The human factor leaders are missing
There’s another layer to this that often gets overlooked. Leaders are investing heavily in AI tools, but many are underestimating the human side of adoption. We’ve seen organisations where the biggest barrier isn’t capability, it’s behaviour - people say they don’t have time to use AI. But when you dig deeper, it’s not time. It’s permission. They don’t feel allowed to experiment, change how they work or step outside established processes.
Without that cultural shift, even the best tools won’t deliver value.
Why “AI-Native” Is often misunderstood
There’s a growing ambition for organisations to become “AI-native.” But in many cases, that ambition is poorly defined. Being AI-native isn’t giving employees access to tools, running isolated experiments or layering AI onto existing processes.
It’s about rethinking how decisions are made, how systems connect and how work actually flows through an organisation. And that requires both technical and behavioural change.
What actually drives productivity
If AI is going to deliver real productivity gains, it won’t come from volume. It will come from removing steps, connecting systems, enabling better decisions and embedding intelligence into workflows. That’s a much harder problem to solve. But it’s also where the real opportunity sits.
The bottom line
AI doesn’t automatically make organisations more productive. It amplifies how they already operate. If your systems are fragmented, it creates more noise. If your processes are unclear, it accelerates confusion.
But if your foundations are strong - clear thinking, connected systems, focused execution - then AI becomes a genuine multiplier. And that’s the difference between experimentation and transformation.
