Microsoft AI Tour - London 2025

As you may have seen yesterday, I was fortunate enough to be able to attend the AI Tour at the ExCel in London yesterday. As much as I was tempted to share yesterday, I wanted to sit back and reflect on what I heard, so I can share back in a meaningful way to you lovely folk.

So travel war stories aside due to the line fault between Reading and Maidenhead (and there were many of them with folk similarly keen to not miss the event!) - my first session was "The next steps to adopting AI in Public Sector) - having spent a lot of time working in this space hearing the ground truth from Andy from Somerset CC and Nicola from HMRC was super enlightening. I loved Andy's statement that their AI journey has not necessarily been solely about ROI, but about ROE - Return on Employee, be it from time savings in creating care plans, saving the time of 4hrs per day when reviewing PDF CV's or just saving enough time for staff to spend a 15 minute break together - all of these will add up to real meaningful change and impact. Nicola shared the perspective of introducing this capability into an organisation made up of many departments and thousands of staff, all in the context of the ~£700 billion pounds that they are responsible for, which in turns funds other government services. Given their mission, it shouldn't be a surprise that HMRC has been using AI for many years now and one of the key learnings Nicola shared was don't look for the technical solution, as you will get technical solutions. Instead reach out, ask the folk on the ground what their day-day challenges are and then see what todays AI capabilities can do to address these challenges. I particularly loved the quote - "security is not a governance issue, it's a risk conversation" and I wholeheartedly agree with this, all too often we see these 'governance councils' putting blockers up which ultimately reduces the pace of innovation. I also loved the point during the discussion on responsible AI and Nicola pivoted this to "actually, this is just responsible innovation" - hearing that out aloud made something click in my brain.. Hopefully it does for all also!

After this session is was then time for the keynote. I was super impressed with Darren Hardman at the outset, a lot of what was said resonated with me. The specially given my long standing work in the public sector seeing the impact AI has had on real people, whether it was 'cutting digital drudgery', improving the government 'Redbox' into a digital briefing solution and other exemplars, it's great to see the tangible impact here - though equally, a latent warning that the heritage we have in the UK here of innovation is not enough - we need to thrive in this AI era. This segued into a discussion on a new paper that was shared - Agents of change - you can find that at http://aka.ms/agentsofchange.

We then shifted into Judson's part of the keynote, I do enjoy listening to him speak and hearing first-hand the pivot from what last year was 'here are some AI principles you should be working to' through to 'here are actionable steps to make the most of AI in your organisation' was really refreshing. He also shared a super interesting datapoint, Microsoft are looking to continue with double digit growth, with zero headcount increase - copilot, agentic AI and whatever comes will of course be at the core of this and in one of the later sessions you got the sense of how this is being driven, as they shared that there are daily reports of (gen)AI use in MS across the entirety of its population and if folk are not using it for at least 50% of their workday then this is flagged up(!). Judson also shared that GitHub Copilot has also written approximately a third of the code for all the other Copilots - this is super interesting again, as there is a lot of debate in the broader community right now about how 'good/fit for purpose' code provided by GenAI is - This is a clear indicator that if MS can make it work for their own products, then for sure the rest of us can - and the beauty here is, we have access to exactly the same stack that the Microsoft folk do, as it's the same platform!

As the sessions progressed it came through loud and clear that for AI to deliver, it has to have exec ownership & accountability. The rise of a 'Chief AI Officer' is a clear signpost to this it also became clear that whilst many of us are 'learn it all' sort of folk, the challenged placed upon us to remain current is significant right now - I loved how Andy Beatman described this in one of the latter sessions of the day, by saying "keeping current with AI is like working in Dog years, so one year of AI feels like four odd years in reality!". And on that track, Andy's session was one of the best sessions for me, clearly grounded in experience and great analogies, plus the Peppa pig wakeup at the beginning 😂 - it was a super, super session!

One final thing before I close out this article - Andy very neatly captured another common theme that I heard throughout the sessions in the day. In this era, "Accessibility is about creating inclusive experiences for all". We are rapidly moving to a place where you will be able to talk your native language, and have real-time voice translation, using your own unique mannerisms & tone. This not only will help those with both visible and non-visible disabilities, but will also bring our world, our global village together - and really.. what better endeavour is there than that?!

And that's it'! as I sign off, I'd love to know what resonated with you if you attended, and even if not, your thoughts on the above.