Steve Poole

Exciting news: over £59 million announced today to support the Government’s new compute roadmap. Our investment will support the career development of researchers and innovators, enabling UK-wide capability and access to digital infrastructure. They include: 🔶 £9 million for five Digital Research Technical Professional (DRTP) Skills NetworkPlus awards to Universities of Warwick, Southampton and Edinburgh, University College London and Imperial College London to boost technical skills and bring together different computing disciplines. 🔶 £6 million for Universities of Surrey and Durham to enhance technical professional skills to match those found across the globe. 🔶 £1.6 million for University of Edinburgh to share best practice and knowledge exchange around the future of supercomputing. By strengthening the UK’s digital research capabilities, we’re unlocking potential for major breakthroughs in areas like healthcare, energy, and public services. From smarter devices and more accurate health diagnostics, to improved prediction and mitigation of the effects of climate change. Professor Charlotte Deane, Executive Chair of EPSRC and UKRI Digital Infrastructure Champion, said: “This new roadmap contains many measures to take digital infrastructure to the next level across the UKRI, bringing many benefits to businesses, communities and families everywhere. “By harnessing advanced computing power, there is the potential to drive progress in many areas of life, such as health, energy, transport and public services. “Our vision is for researchers and innovators to be able to access and exploit the best digital infrastructure for the benefit of lives and livelihoods across the UK.” Today we’re also delighted to see the UK’s most powerful AI supercomputer is now online ✨ Isambard-AI, based at the University of Bristol, is set to give UK researchers and industries access to world-class computing power to train AI models and process complex calculations at lightning speed. It's already being used by University College London researchers to analyse MRI scans so we can identify cancer earlier and provide patients with personalised treatment plans - this tool could soon be rolled out across the NHS. Together with the “Dawn” supercomputer at the University of Cambridge, Isambard-AI forms the £350 million AI Research Resource, and will see the UK's compute capacity increase to 500 AI exaFLOP - the equivalent of one billion people spending 15,844 years doing what the full AIRR will do in one second. Early users of the system have been able to conduct research that's currently not possible to do on conventional systems, highlighting how this infrastructure will unlock innovations that will improve lives and drive economic growth across Britain. Head to our website for the full story.

240

4 Comments

To kickstart a growing community of Federated Developers, the Data and Navy Applications team (DNA) ran a two-day Hackathon event in Portsmouth, one day to learn, one day to build. The mission: empower UK Crown Servants and Royal Navy personnel to tackle digital challenges using coding skills. With guidance from digital experts, and judges from the RN, Civil Service, and Microsoft, teams built applications that address real operational pain points. The result? A glimpse into the future of Navy-led digital innovation.

197

6 Comments

The Ministry of Justice UK has just released a new online directory of open standards, manuals, products, design systems, and service patterns. Super useful for anyone working in product and technology in the public sector (and outside of it). https://lnkd.in/eM69f_2N

50

2 Comments

We're proud to be partnering with the Department for Science, Innovation and Technology to launch the AI Engineering Lab - a ground breaking initiative supporting the adoption of AI coding assistants across UK public sector software engineering teams.   What makes this programme different is our focus on people, not just technology. Alongside tools and licences, we're working with teams to support the change needed for AI to be genuinely adopted and deliver real impact.   Through structured coaching, tailored training, and peer-to-peer champion networks, we're helping departments get meaningful value from AI coding assistants while building capability that lasts. Because technology only delivers value when people are supported to adopt new ways of working.   Find out more: https://lnkd.in/eKngvYbN   Claudia Chiurlia, Dimitris Perdikou, Simon King (FBCS), Sarah Handyside, Bharath Vadhoola, Ali Bovis, Steve W., Emma Low, Cleo McCormack, Sam B., John O'Neill, 💡 Maro R., Colm Lynch, Anthony Stuart 

201

A lot of migrations are an absolute car crash because people don’t check their mirrors. And no I am not talking about the political issues in the UK when I say migration😂 The sensitivity around migrations can sometimes be overlooked. Yes there is this new nice tool that might do the job faster or have better integrations to enhance the developer experience. But how do you get everyone onboard… How do you minimise downtime… How do you get key stakeholders onboard for the new initiative. Being able to justify a migration is the first step. Next is effectively executing it without causing regret. Sometimes the grass looks greener. The worst conclusion is now experiencing new problems and having engineers say… “Things were not perfect before but atleast it worked” Remember to evaluate all the important factors: 1) Cost 2) Compliance 3) Operational overhead The list goes on and on. Or as us Gen Z individuals like to say. Remember to check your mirrors.

151

5 Comments

Great news as CodeBase takes on the recently closed CodeClan, with Scotland’s digital skills academy set to relaunch in early 2026 with new courses including software engineering for first timers, and applied agentic AI for experienced engineers 🎉 The hybrid, project-based programmes will let participants upskill while working, combining real-world projects, industry mentoring, and flexible learning. “We don’t just want to teach code, we want to build futures.” said Stephen Coleman, CEO of CodeBase. This relaunch marks a major step toward closing Scotland’s digital skills gap and preparing talent for tomorrow’s tech economy 💻🤖 For the full article 👉 https://ow.ly/ALBb50XgYiS #DigitalSkills #AI

40

3 Comments

Richard Sutton’s “Bitter Lesson” is completely changing growth rates and funding this AI cycle. a16z General Partner Martin Casado explores: What does this mean in the long run? Does this spell the end of traditional software? Does this mean we will see a diminishing need for software engineers? Find out: https://lnkd.in/dm6AE7Qk

199

25 Comments

Another week, more AI action - this time the UK Compute Roadmap. There's loads of good stuff in this - its genuinely worth checking out but here's what I'd draw your attention to: 1) The Roadmap sets out how the UK Gov will spend up to £2 billion to build out the UK's public compute. 2) A big chuck of this compute will be used to deliver on national priorities through the UK’s AI Research Resource (AIRR). This gives the UK Government compute to allocate to specific areas of research and public missions. This is part of the journey to x20 the amount of public compute in the next 5 years as set out in Matt Clifford's AI Opportunities Action Plan. 3) Large scale AI infrastructure will be delivered via AI Growth Zones (AIGZs) - a core group of nationally significant sites will be set up, each capable of serving at least 500MW of demand by 2030, with at least one scaling to more than 1GW 4) The Sovereign AI Unit (where I work) - has £500m to make things happen. In a Government context this is huge. In an industry context this is tiny. That means we have to work on novel areas. We're doing just that, more below 5) Sovereign AI has a reserved allocation within the AIRR - this means we have live compute capacity to support our goals. At the end of June we launched a 1 week call (now closed!) for highly ambitious projects across a range of priority areas. We got some really exciting proposals and I hope we can say more on what we are taking forward soon. 6) Compute will be a priority area for Sovereign AI - this means supporting R&D into new computing paradigms. If you have ground-breaking ideas we want to know There's real momentum - roll on the next series of announcements we're able to get over the line. Links to relevant info: Sovereign AI AIRR call for researchers: https://lnkd.in/e5BW_3Nh Compute Roadmap: https://lnkd.in/eeB26M8K Compute Roadmap Press release: https://lnkd.in/eG_U3R8p

131

10 Comments

We've all seen the headlines about the government wanting to get people into work. A thriving economy relies on a skilled and active workforce, yet PwC UK's latest research shows that one in ten adults are thinking about leaving the workforce.    Business leaders are concerned, too—81% of those surveyed reported that productivity and financial performance have been negatively affected by economic inactivity. What’s driving this trend, and how can businesses and policymakers work together to tackle it? Read the report to find out more https://lnkd.in/eG3Wusnv

9

🚀 Why should the public sector use AI coding assistants? The Engineering Excellence team at the Government Digital Service has run the UK’s first cross-government trial of AI coding assistants (AICAs), including GitHub Copilot and Gemini Code Assist, to understand their real impact. 👉 1,900 licences distributed across 50+ organisations 👉 Average 56 minutes saved per developer, per day 👉 Strong positive feedback, even from veteran programmers The evidence shows that AICAs can boost productivity, reduce duplication of effort, and support smarter procurement decisions across government. 📖 Read the full write-up, including results, methodology, and next steps: https://lnkd.in/evPW5fYG #DigitalGovernment #AI #PublicSectorInnovation #GovTech #GovernmentDigitalAndData

49

4 Comments

Brian Randell born in Cardiff, Wales occupies a rare space in the history of computing: he is both a builder and a chronicler of the field. A pioneer in dependable systems and fault tolerance, Randell has also been one of the most rigorous historians of computing, particularly the formative years of the discipline. His contributions stretch from the design of reliable computing systems to a foundational role in shaping our collective memory of early computer development. Randell studied mathematics and then computing at the University of Cambridge, before joining English Electric in the 1960s—one of the few companies in Britain seriously pursuing computer development at the time. It was there that he helped design the Whetstone ALGOL compiler and became deeply involved with compiler construction and system software. His work on ALGOL introduced him early to the intricacies of language design, parsing, and the challenges of writing robust and efficient software on constrained hardware. His most influential technical work, however, came after he joined Newcastle University, where he led groundbreaking research in fault-tolerant computing. At a time when most assumed that hardware and software failures were nuisances to be avoided, Randell treated them as central design considerations. He introduced recovery blocks and advocated systematic approaches to error detection and correction—laying a foundation that still informs high-reliability systems today, from avionics to distributed cloud architectures. Randell’s 1975 paper with Randell and Rushby, "Design Diversity and the Decomposition of Software Faults," was a landmark in the study of software reliability. He understood early that software engineering wasn’t just about producing functionality—it was about engineering for resilience in the face of the unpredictable. His work contributed directly to the birth of dependable computing as a subfield, influencing generations of researchers in distributed systems, redundancy, and formal verification. In the late 1970s, he turned his attention to documenting and interpreting the early history of computing, particularly the British contribution. His 1973 edited volume The Origins of Digital Computers: Selected Papers is still a core reference. Randell tirelessly tracked down primary sources, corresponded with pioneers like Turing’s colleagues at the NPL, and debunked myths—most notably about the early machines built in Bletchley Park and Manchester. He played a crucial role in bringing Alan Turing’s legacy to broader awareness and helped secure the recognition of Tommy Flowers and the Colossus team, whose wartime work had been long buried in secrecy. His historical writings are not nostalgic tributes—they are rigorous, source-based, and deeply analytical, reflecting a mind that understands both the technical and the human aspects of invention.

48

8 Comments