Post

At the first session of the #ENGINE meeting at #HarwellCampus. Looking like a really great meeting with all the groups on campus present. Held at the new #CO:LABs building which is co-founded by the @rosfrankinst.bsky.social. The introductory keynote is by #LordDavidWilletts.

Second session at the #ENGINE meeting, all about the developments of the 3 national facilities on site, @diamondlightsource.bsky.social @isisneutronmuon.bsky.social and the #CentralLaserFacility.

Sorry @clf-stfc.bsky.social I managed to not find you when I was posting!

The next talk is by #DameNicolaBlackwood about what's next for UK life sciences. Especially at the #HarwellCampus highlighting how the @rosfrankinst.bsky.social is perfectly situated to make the most of the capabilities available here.

Some really good points made by many of the speakers today about challenges around not just large quantities of data, but that it needs to be of exceptionally high quality to be useful for AI and other methods to not just reinforce our biases. It's great to hear these ideas gaining traction widely.

Really interesting and inspirational talk by Dame Nicola Blackwood. Her vision for transforming healthcare over the next decade is very compelling and achievable with the right focus. Looking forward to working towards our small part of this!

Next up is #ProfessorCharlotteDean talking about AI for science, and what's next. Highlighting the complexity, but now alignment of government policy around AI and science strategy. Also thinking about #ResponcibleAI and how we can use these new tools without negative impact.

I like the idea of Human-on-the-loop as a view to where a fully internal optimisation loop can be created, but where the human effectively modifies the loop itself to make it work as well as it can.

Just finished my session on AI for predictive biology which was a lot of fun. Next up I wanted to learn more about the @openbind.bsky.social project with #FrankVonDelft, @fergusimrie.bsky.social and @elizabethshotton.bsky.social. Really nice synergy between experimental and AI research.

Really like the description of @fergusimrie.bsky.social around how much data is needed is different to get to a generalised model versus an interpolated model.