4 July 2025
Empowering the next wave of AI-powered science
The Encode Fellowship – run by our Activation Partner, Pillar VC – is bringing top AI talent into scientific labs to accelerate breakthroughs from medicine to climate change.
For years, a bold idea captivated a mathematician-turned-ML engineer: that AI held the key to cracking ageing and cancer. He spent four years at a tech conglomerate working on a successful biomedical-AI project that clinicians initially called "impossible," yet his research only deepened his conviction, and left him hungry to push the science further. Now, through the Encode Fellowship – made possible through our activation partnership with Pillar VC – he’s able to pursue his ambition full-time at the Francis Crick Institute, advancing his own goals while potentially deepening our understanding of these diseases.
His decision represents something profound happening in UK science – and helps explain why the government announced additional funding this month to double down on the Encode Fellowship's potential.
The next AI revolution is science
"The biggest lever we have on the speed of science is AI," says Tony Kulesa, Partner at Pillar VC. Already, AI systems are modelling protein structures and simulating planetary atmospheres in a fraction of the time it once took. But behind these systems are brilliant scientists and engineers with a passion for solving complex problems.
As part of our Activation Partners initiative, we're working with early-stage VC firm Pillar VC to pair world-class AI experts with top researchers — injecting entrepreneurial energy into science to unlock new breakthroughs.
By doing so, the ambition is to bridge a growing divide that could hinder progress. As our Chief Product Officer, Pippy James explains, “Much of the world's top AI talent isn't currently focused on fundamental, hard science problems." ”They feel cut off from the domains where they could have the greatest impact,” adds Leah Morris, Executive Director of Encode at Pillar VC.
On one hand, scientific research institutions lack access to AI talent (nearly 70% of new AI PhDs are drawn to industry where the talent pipeline flows overwhelmingly away from fundamental research). On the other hand, the most talented AI minds often feel they could unleash more potential if they applied their skills elsewhere. This means the power to revolutionise science exists – but we need to build the bridge between these two worlds.
A shared vision for impact
The Encode Fellowship represents a counterintuitive solution: placing world-class AI researchers directly into academic labs working on challenges aligned with ARIA's opportunity spaces.
Engaging with potential fellows before designing the programme revealed a clear, shared ambition. They were drawn by the intellectual satisfaction of building models to tackle the most important questions in science. This desire for transformative impact was matched by a need for proximity to the real-world instruments of discovery – from fMRI scanners to wind tunnels – expensive and specialised equipment that largely exists only in academic research settings, and is essential for generating fresh data.
Build the opportunity and the talent will come
When applications opened in March, the Encode team received more than 600 applications from researchers at leading companies like Anthropic, Meta, Ginkgo, Isomorphic Labs, and Google DeepMind, and top global universities including Harvard, IIT, Berkeley, ETHZ, and MIT. Of these, 40% came from the UK, 25% from the United States, and the rest from 54 other countries – all incredibly accomplished and published researchers wanting to build in the UK, and looking for alternative choices to traditional academic career paths.
The demand validated the hypothesis behind the Fellowship: top researchers want to work on hard science problems, there just haven't been enough opportunities for them to do so.
Consider the mathematician who chose the Crick – already convinced that AI could tackle ageing and cancer, he needed the freedom and resources to pursue this full-time. Or take another final-year PhD student using AI for coral-reef conservation. Over five years, they've conducted 300+ scientific dives and now want to unlock petabyte-scale 3D twins of reefs with Google-like natural-language search capabilities.
As Leah explains, the Fellowship provides what these researchers need the most: "one year of backing, compute, world-class lab access, and crucially, deep-tech mentorship and a venture-focused network to build a viable product that has real-world impact.”
Gathering momentum
The UK has long had the essential ingredients to attract global AI talent: world-class institutions, vibrant science clusters, and entrepreneurial researchers. But programmes like Encode are being helped by both new thinking and new investment in the UK.
ARIA's mandate to back unconventional ideas and establish new partnerships, combined with the UK government’s £14bn AI Opportunities Plan to develop new talent-friendly pathways for global researchers to embed themselves in UK labs, is helping ambitious programmes like these make the most of the UK's world-class research network.
Policymakers have spotted the opportunity. At London Tech Week, DSIT doubled down on Encode, announcing an additional £5m in funding — a signal of growing confidence in the UK’s potential to lead in AI for science, and in the Fellowship’s ability to help catalyse that progress.
"The Encode Fellowship is helping us go a step further, putting world-class research talent into labs to foster deeper collaboration and bring truly game-changing breakthroughs to life faster than ever before," said Minister for AI, Feryal Clark. "Encode will help to strengthen our position as a leading destination for AI talent, and build up our sovereign AI capabilities as we put the technology to work to deliver on our Plan for Change.”
The collaboration between world-class AI talent and leading academic institutions represents exactly the kind of ecosystem building that ARIA was designed to catalyse. "It takes two sides to make the Fellowship work," notes Pippy, "and it would never be possible without the willingness of outstanding UK research labs to host these individuals."