Five senses AI: machines now see and hear, but none can smell, touch or taste. What ARC-AGI-3 and ARIA reveal about the gap between AI and superintelligence.

Redundancy feels like a permanent scar. Nobody else can see it. And in our own study, the weakest radiologist combined with AI beat the strongest working alone.

Unsolicited VC calls told me the business was scaling. The deal we actually took was from a hospital network, and the engine had nothing to run on.

The government has accepted all 44 recommendations on regulating AI in healthcare. Recommendation 16 is the only one that changes what a UK approval is worth abroad, and why my advice to British SMEs is still: build evidence here, go FDA first.

How to Build a Start-up, Part 6, Author: Simon Rasalingham, 2026 The product you launch with is rarely the one that makes the money. Ours took off a decade later, on the back of a single guideline. MEDICA started as a chest x-ray reporting business. The logic was easy to understand. As x-rays moved from pieces of film, which could be hidden in a cupboard, to images sitting on a system, radiology managers across the NHS suddenly discovered how many chest x-rays had never been

Why Set Up a Company in Britain? Two asks from a British entrepreneur who wants to. Neither requires rejoining the EU, neither is a subsidy. Author: Simon Rasalingham, 2026 Foreword A founder starting a frontier life sciences AI technology company today has no rational reason to do it in Britain. I am asking for two changes in this Budget, neither of which requires us to rejoin the EU. The prize for getting it right is a trillion dollars of economic impact. I have been involv

How to Build a Start-up, Part 5, Author: Simon Rasalingham, 2026 Most company names get decided in a pub. My first permanent job after university was at Guinness, surrounded by some of the best brand people in the world, working on something older than the United States of America. Guinness was created in 1759 and is one of the oldest and most recognised brands in the world. The purpose and identity of Guinness was drilled into every single employee, from the person installi

Author: Simon Rasalingham, 2026 On my first visit to the FDA, driving up to the entrance, I wondered why there were men with semi-automatic weapons at the gate. Then I took my shoes off to get through security, to attend a meeting about clearing an algorithm. The meeting was my team on one side and two rows of FDA people on the other. For a first meeting. That is where most British founders' regulatory strategy gets decided. It is very intimidating, and it can put you off for

How to Build a Start-up, Part 4: The Financials, Author: Simon Rasalingham, 2026 You have a customer, a solution and a business plan. The next step is the financials, and most founders build the wrong spreadsheet first. They start with the five-year model: revenue curves, a hiring plan, a profit-and-loss forecast that turns positive in year three. It looks rigorous, and it answers a question nobody is asking yet. The question that matters in year one is simpler: how much cash

Author: Simon Rasalingham, 2026 A new benchmark paper gives us, for the first time, a single number to compare quantum computers across platforms. The number is not flattering — and the reason it isn't flattering is more interesting than the number. The headline finding, stated plainly: the best quantum computers on Earth are not losing the race to classical machines. They have not yet reached the starting line. Everything below is the evidence for that, and what it should do

44 Recommendations, and Still Nobody Owns the Liability Author: Simon Rasalingham, 2026 The National Commission into the Regulation of AI in Healthcare has reported. I gave evidence to it in January. The result is 44 recommendations across 119 pages, and it is serious work. Across those 119 pages: Industrial strategy — no mentions. Reimbursement — no mentions. Sovereign AI — one mention, and that was about the risk of relying on foundation models owned outside the UK. And the

How to Build a Start-up, Part 3, Author: Simon Rasalingham, 2026 I wrote two business plans for MEDICA. The first was a three-pronged attack on revolutionising radiology. The second four year later was focused on one. The focused one built the company. The first plan came when I was starting MEDICA. I had spent time at Guinness World Records, worked with some talented consultants from Bain, and had an MBA. I came away with a head full of good frameworks, which I used to help

How to Build a Start-up, Part 2, Author: Simon Rasalingham, 2026 I have always been a curious person, sometimes too curious. I remember walking into a server room at an NHS hospital just as I was starting MEDICA. A wall of servers, humming. I asked the IT manager how much it had cost. "Two million pounds," he said proudly. I had just come from Guinness, where I had built out the whole technology infrastructure on the latest Windows NT. I knew roughly what that capacity cost a

How to Build a Start-up, Part 1, Author: Simon Rasalingham, 2026 I won my first tender in a meeting room at Whipps Cross University Hospital. It came down to one question about costs. I was down to the last two. They told me the other supplier had said that integrating with the hospital's local radiology system would cost £50,000. They asked how I planned to do it. If my cost was the same, then the two bids were the same, and they had more pedigree. I was just an idea and a s