Breaking bread over breaking science
Announcing our first cohort of Innovator Circles: high-trust peer groups to shape ideas at the frontier of science and tech
Why peer groups matter for technological breakthroughs
Highly ambitious, technical breakthroughs rarely arrive fully formed. They’re often born out of half-articulated thoughts – whispers in the university library, debates over lunch, or musings on the walk home.
ARIA’s Innovator Circles initiative was designed to support these environments: high-trust, self-organised technical peer groups where people can test ideas early, raise each other’s ambition, and pursue ideas that could underpin the next world-changing breakthroughs.
We tested this hypothesis last year through a focused experiment, supporting a small cohort of individuals to build their own peer groups. The results, shared here, showed us the potential for cultivating ambition more broadly — creating spaces for people who may not feel they ‘fit’ with traditional scientific institutions. By creating informal environments where people can contribute without fully formed ideas or established credentials, we saw how our experimental cohort opened up exactly those spaces. Furthermore, by targeting an earlier stage of the pipeline to traditional funding models, such as grants, this initiative opened up new pathways into science and technology, diversifying both ideas and the pool of people working to generate them.
To continue building on this momentum, we doubled down with a call for individuals to build their own Innovator Circles across the UK.
We received close to 130 applications from around the country. ARIA’s offer was to support these groups with up to £2,000 a month to set up their Circles, alongside access to ecosystem partners, and, importantly, a ‘social licence’ to build high-trust, high-ambition peer groups. The quality of these applications reflected the breadth and diversity of latent talent and ideas waiting to be activated with the right kind of support.
From this pool, we selected 10 Leads to establish Innovator Circles in topics ranging from exploring the immune system as an operating system, to compositional growth of neural networks. What unites these Leads is their willingness to challenge prevailing paradigms, combined with a vision towards building real-world peer groups that could catalyse work on new approaches to expanding the frontier.
Two months into the initiative, we’re beginning to see that these environments often recreate something that has historically been critical to innovation. As Boyuan Xiao, Lead of Hacking Biomechanics Innovator Circle, shares:
“So much innovation comes from informal encounters with those working in other fields, and memoirs from places like Bell Labs are filled with odes to the seminars, corridors, and canteens which brought scientists together.”
What the Circles are exploring
Afiq Hatta - The Independent Science Society
Many capable and technically ambitious individuals lie outside of academia. The Independent Science Society aims to figure out how effective scientific exploration can be done outside of conventional environments.
Alexander Ainscough + Inga Van den Bossche - ImmunOS
What if we treated the immune system as a programmable biological operating system? By viewing immunology through the lens of computation and AI, this group aims to unlock silos and discover how to ‘upgrade’ our biological intelligence. To get in touch, reach out to
Get involved: research@immunos.bio
Boyuan Xiao - Hacking Biomechanics
This group is democratizing high-end biomechanics research. Using 3D printed parts and open-source software, they’re exploring ways to maintain human mobility through accessible prostheses, exosuits, and other devices that help us to move freely, especially in the face of aging-related conditions like muscular atrophy and arthritis.
Charles Grellois - Compositional Growth The Algebra of Complex Neural Structures
Moving past the fixed architecture paradigm, this Circle is investigating neural systems that dynamically expand their structure based on demand. It’s a ‘growth-by-need’ approach that bridges mathematics and neurology to create more energy-efficient, explainable AI.
Edward Miller - Computational Models of Human Experience
This group is defining a field that doesn’t exist yet: the scientific foundations of human foresight. They are asking what it would take to genuinely model and predict human behavior at a resolution high enough to guide the most consequential decisions of our future.
Isabelle Zane - Unconventional Biology
By focusing on the “edge-cases” of life — extremophiles, synthetic cells, and xenobiology — this Circle seeks to redefine what constitutes life itself, to find new therapeutic and industrial inspirations.
Get involved: hello@unconventionalbio.com
Ivy Li + Içvara Aor - Rethink Matters
Biomaterials cannot succeed on technical promise alone. Their future depends on whether they can connect across fragmented disciplines, survive scale-up, and navigate regulation and governance. This Circle explores those often-hidden challenges to understand why breakthroughs stall and what it takes to turn them into viable and fair next-generation materials.
Nandini Shiralkar - AIxBio & AIxSocial-Science
AI is not a “normal” technology; it is a general-purpose capability shifting the structure of power and knowledge. This Circle is convening domain experts and AI researchers to map the second-order risks in biosecurity and social systems that both groups currently miss in isolation.
Get involved: AlxBiosecurity, AlxSocialSciences
Neil Shevlin + Sam Enright - The Fitzwilliam AI Circle
While there is a huge amount of focus on easily verifiable LLM tasks, this group is asking if AI can scale into domains where rewards are messy and “scientific taste” is the only compass. This includes exploring open-ended scientific discovery in fields as complex as mineral exploration, and whether we can build AI systems capable of open-ended scientific discovery – focusing on provisional objectives, evolving research directions, and mechanisms for developing research taste beyond fixed, verifiable tasks.
Rakesh Arul - Quantum Utility
Most quantum research happens in precisely controlled, ultra-cold labs. This Circle is pushing quantum science into the “wet, noisy, and room temperature” environments of the real world — targeting immediate advantages in molecular and biological contexts.
What Innovator Circles look like in practice
Innovator Circles are deliberately lightweight and flexible. Rather than prescribing a format, ARIA supports individuals to convene groups in a way that suits them.
In practice, this has resulted in a wide range of formats: technical reading groups, hands-on demos, informal dinners, co-working sessions, and structured debates.
Across the cohort, Circles span a wide range of technical areas, but are united by a focus on exploring ideas while they are still incomplete, and open to challenge. Nandini Shiralkar, Lead of Alx Innovator Circle, says:
“I wanted to create a space where those people can think candidly, disagree seriously, and build sharper models of the world frontier AI could create. I’m hopeful that this could change what circle members decide to work on, and more importantly, what they believe is even possible to build in the first place,”
Early signs of impact
The Innovator Circles are not designed to produce polished outputs. Instead, we’re exploring early signs of impact: new collaborations, shifts in ambition, and the emergence of ideas that might not otherwise have formed. Several patterns are already visible.
The value of building new collaborations across disciplines has emerged as a common theme. For the Quantum Utility Circle, the permission to bring together individuals of different backgrounds has encouraged collaborations between traditionally separate areas. The Circle is regularly bringing together chemists, spectroscopists, biomedical scientists, and industry engineers to discuss the applications of quantum science in the real world, away from precisely controlled laboratory settings.
Circles are also moving their discussions into real-world action as well. In the Hacking Biomechanics Innovator Circle, knowledge sharing between participants from different fields has led to detailed plans for trying out new servo motors and experiments with electromyography, for improving the actuation and motion prediction in a lower limb prosthesis.
Others are exploring entirely different models of how science can be advanced. The Independent Science Society Circle is focussed on how scientific discoveries can be made outside of conventional academic environments to increase the diversity and breadth of scientists, and in turn increase the ideas we can explore and discover.
These are early signals, but they point to a consistent pattern: when the right conditions are created, people begin to connect, think differently, and act on new ideas. At a national level, this early-stage, cross-disciplinary brainstorming is critical to the UK’s ability to generate the next generation of scientific and technological breakthroughs.
What’s next
The Innovator Circles initiative is designed to run for 10 months - allowing the groups to build self-sustaining momentum from scratch.
Over the coming months, the Circles will experiment with different formats, including reading retreats, conferences, hackathons and research agendas. They will deepen their intellectual foundations on the topics they’re exploring as well as grow their groups further before producing outputs such as white papers, code repos, and publications to share their findings with wider ecosystems.
We invite those with expertise in the Circle areas to engage with the groups through attending meetups, reading their updates/newsletters or sharing their own work with the relevant Circles.