The fourth Accelerating AI Conference arrived at the Ridgeway Centre in Milton Keynes on Tuesday with a message that its organisers have been delivering, free of charge, for four consecutive years: the gap between businesses that understand artificial intelligence and those that do not is widening, and the window for catching up is narrowing.
The event, put together by Andy Paul of Fliweel.tech, Matthew Rigby-White of Qoob and Lionel Naidoo of Milton Keynes-based Dragon Information Systems, drew one of its largest audiences yet. Every speaker gave their time without fee. “There is a moral obligation to share what we know,” said Andy.
Microsoft opened proceedings with a number that framed everything that followed. The company currently has 450,000 AI agents running within its own operations, most of them built not by engineers but by staff across the business who identified a problem and resolved it. Louis Nicholson of Microsoft was unambiguous: the question of whether artificial intelligence matters to business has been answered. Organisations that have not yet accepted that are already behind.
Lionel Naidoo of Dragon Information Systems made the case that smaller businesses, often anxious about their ability to compete with larger and better-resourced rivals, may have misread their position. The absence of legacy infrastructure and institutional inertia is, he argued, a structural advantage. “The bottleneck isn’t the technology,” Lionel said. “It’s organisational readiness. And that’s where smaller businesses have the edge.”
Matthew Rigby-White of Qoob addressed the gap between perception and reality in AI adoption, warning that issuing staff with software licences is not a strategy. His agency recently built a fully operational client website in three days using AI tools — a project that would previously have occupied his team for weeks. The efficiency gains are real, he said, but so is the risk of losing the human quality that distinguishes good work from automated output.
Milton Keynes itself featured prominently in the day’s argument. Lisa Beckett, Head of Data and Customer Insight at the city council, described an AI transcription programme in adult social care that has recovered approximately 60 hours of administrative time every week, redirecting it to frontline staff. Ian Pulford of Smart City Consultancy noted that Milton Keynes was the first city in the United Kingdom to permit Starship delivery robots on public streets, and that autonomous vehicle trials continue there today. The city does not need to make the case for its technology credentials. The evidence is already on its streets.
Robert Simpson, Director of Business Development at Zoho, whose UK headquarters is in Milton Keynes, turned the conversation to what he sees as the real obstacle facing most organisations. Businesses, he said, tend to frame AI as a question of intelligence. The real constraint is elsewhere. “The problem inside most organisations is not a lack of intelligence,” Robert said. “It’s a lack of capacity.” In five years, the organisations worth studying will not be those that produced the most compelling demonstrations of what AI can do. They will be those that used it to make work flow better and decisions more reliable.
The panel session, moderated by Usman Ikhlaq of techUK, closed with a figure that unsettled the room. Twelve months ago, artificial intelligence was writing between four and five per cent of the code that feeds into the next generation of AI models. That proportion has since risen to 70 per cent. The capability of these systems is doubling every four months, down from a six-month cycle a year ago, and the interval continues to compress. “The cost of intelligence is dropping like a rock,” said Andy. “It is becoming democratised. For the first time ever, young people have access to intelligence that no generation before them has ever had.”
The fifth Accelerating AI Conference takes place on 13 October 2026 as part of MK Tech Week. Details are available at acceleratingai.co.uk or book direct at: https://luma.com/uxod18a9
