Is UK plc ready for the super-agent era?
Right now it feels like the AI conversation oscillates weekly between declarations of a new industrial revolution and predictions of an overhyped bubble.
Both are probably true. AI may be simultaneously massively overblown in the short term and wildly underestimated in the long term.
This pivotal moment demands two things from UK tech leaders: avoiding the danger of doing nothing until we get certainty, and resisting the hype that distracts us from our core purpose. The 'how' of our business is about to be reinvented, but the 'what '— our unique expertise — has never been more important.
As the leader of a company that’s powered by data, serving tens of millions of customers, I have an obligation to steward my organisation through this shift. While no one — not even the largest LLMs or strategy consultants — knows the exact future, one thing I do know is we have to prepare.
With the advent of AI, the tools at our disposal are rapidly improving and it’s vital we harness them to provide better customer experiences — just like our forebearers did with the internet, telephone and printing press.
So where will the space be for UK businesses in this mix?
The agentic future: spheres of influence
As usage of Gemini, ChatGPT and others increases, it’s tempting to imagine they'll end up monopolising the agentic evolution of LLM. But I don’t believe we’ll see a consolidation to singular, all-powerful agents.
Even if the LLM giants become the new front-doors, the need for an ecosystem of specialist businesses will remain.
I think we’ll see the emergence of several super-agents, each with a distinct sphere of influence — think travel, health, home management and so on. Underpinning each will be specialised sub-agents, much like the evolution of applets in WeChat and Alipay, to handle more complex tasks assigned by their super-agent.
The new normal will likely be continued competition and innovation between these super-agents, often with regional variations, as they try to push the boundaries of their clustered spheres. Arguably we can see strong signals in this direction, with OpenAI recently announcing selected third-party apps inside ChatGPT.
Why do I believe this?
- Trust limits will prevent monopolies: While some consumers will trust a single entity with everything, most won’t; I might trust my bank's agent with spending advice but I don't want it choosing my mortgage if its incentives are conflicted. Certain brands have built huge consumer trust in specific spheres and regions — they'll have a big role to play building consumer confidence in the next generation of experiences.
- It’s the fastest and most efficient way to get great agentic experiences: Agents need data and context to fulfil a task well — even LLMs will find it more efficient to allocate tasks to high-performing specialist agents over trying to complete every task in isolation.
The outcome will be a need to balance massive AI utility with real life human adoption behaviours, all while ensuring national and global economic ecosystems can still innovate. I’m certain there will be a role for businesses that can earn and maintain consumer trust while using this technology to offer real utility.
How we’re facing into it
So that’s how I see the future — here’s how we’re preparing at RVU:
1. Doubling-down on our value
The interfaces people use to find services will likely change beyond recognition; how customers discover and interact with us is a challenge we must meet head-on. But that doesn’t change our core value. Our deep expertise in our specialised sectors is our anchor.
The challenge isn't whether our expertise will be needed, but how we make it discoverable, integrated and valuable in the new agentic world.
2. Turning our data into a lever
For many companies, data is an unorganised and underutilised liability. Don’t underestimate the importance of well-structured and accessible data.
We’ve worked to get our data house in order, not just for our own benefit, but to allow consumers to leverage their own information for better outcomes. A mountain of data is useless until it’s organised into a powerful lever.
3. Putting the tools in everyone’s hands
The only way to truly understand AI’s capability is to use it. We’re actively encouraging every team to experiment safely to rapidly scale their comfort level. This isn't the responsibility of a single 'AI team,' but an effort to build company-wide literacy.
In doing so we not only better serve the business, but equip our employees with the skills they'll need in their careers in the UK knowledge economy.
Focus on what; experiment with how
The shift to an agentic future isn’t a distant threat, nor is it a guaranteed utopia — it's a current reality that demands strategic action. If UK plc fails to embrace the agentic future, we won’t meet our customers’ growing expectations and we’ll inevitably be left behind.
At this pivotal point, the defining task for every UK CEO is to maintain a laser focus on your 'what' — your unique value — while obsessively experimenting with the AI tools that will redefine your 'how'.
This isn't the moment for panic or passive observation; it’s the moment to ensure you’re ready for the agentic economy — however it unfolds.
