We’re currently recruiting for our 2023 cohort at Conception X, a UK deeptech venture programme for PhD students who are looking to start a deeptech company.
Candidates from STEM backgrounds are prioritized, and every year we publish a list of areas that we think are going to have an outsized impact in the coming year, whether because the market is finally viable or no incumbent is on a trajectory to achieve a position of dominance.
If you are operating in any of these fields or technologies, we strongly encourage you to apply and reach out to me personally if you have any questions — even after the deadline on 27 January. Just like last year, however, please bear in mind this list is by no means exhaustive.
Conception X accepts applications from individuals as well as teams, provided the main applicant is a PhD student. More information in our FAQ.
In no particular order:
Almost 40 years after Engines of Creation, we are finally starting to see the horizon of nanotechnology. The tool stacks are not yet here, and the machines are more chemical than mechanical still, but we have previously invested in this field and want to do more.
Full-stack quantum computing software and applications
At the current rate, it might take two decades before quantum computing becomes mainstream. Nevertheless, we are interested in helping founders who are making progress across the full stack of QC, especially beyond hardware, where all the current generation of companies is concentrated.
Space Economy supply chain
In the span of a few years, the space economy will completely change how certain operations or data transfer happen on Earth. The proliferation of constellations, their associated bandwidth, the ability to perform inference in space — these are all happening. So we want to help entrepreneurs who are deflating costs, adding new capabilities in space, or leveraging existing ones on Earth to achieve what was previously too costly.
Net zero materials and circular economy
Achieving net zero is a moral imperative, but we are looking for startups that can do so while turning a profit, which is necessary for scaling operations. This is a very asymmetrical market, where some novel solutions are entering an overcrowded market, while other problems remain unaddressed.
Biology is becoming a software problem, and we are keen to explore solutions that leverage CRISPR-Cas9 in a variety of settings to improve the quality of life for humanity, both in healthcare and economical terms (such as improving yields in agriculture etc).
Moore’s Law scaling is over, and AI models are getting bigger — with no signs of slowing down. We are looking for a solution to ensure these two issues never collide. There might not be a future for generic purpose microprocessors, but perhaps domain- and problem-specific architectures can bring neural networks everywhere. We are particularly interested in solutions that can merge silicon and software for AI.
Detection of volatile compounds
The advancement of machine learning has been outstanding, with new progress announced almost on a daily basis. But all these have so far relied on purely digital data and the output in software only. Easy to implement, but with narrow impact. We are seeking founders tackling problems of detecting physical compounds or signals and then making predictions based on them. Healthcare, industrial, or simply novel use cases with no clear market yet are welcome.
Transformers, geometry, and graphs
Large language models are making an impact, and we are seeking anything that might augment LLMs or enhance our ability to understand the limitations of AI and how to address their negative effects. Non-Euclidean/geometric ML or graph-based structures to give models a deeper understanding and be able to use operators to manipulate results. This is both an organizational and a technical problem, and we are happy to consider either approach.
Special thanks to John Tang for his input in making this list.