Interview
Share
Tristan's career has involved using AI as a management consultant, trader in investment banks (UBS), portfolio manager in hedge funds (Aspect Capital), research fellow in medicine and as an entrepreneur where he has co-founded/been an early member of several AI-focused tech companies (such as Thought Machine). He has maintained a parallel career in academia, with masters degrees in Manufacturing Engineering (Cambridge) and Computer Science (Sussex), a PhD in Machine Learning in financial markets and commodities pricing /trading (UCL) and a research fellowship on AI in medicine (Imperial).
He loves bringing theoretical, state-of-the-art academic research to life in the real world - whether it's predicting fine wine prices, the location and frequency of ambulance call-outs or diagnosing cardiovascular disease. ChAI, where he has been the CEO since he co-founded the company in 2019, is the perfect expression
of that motivation. It uses AI on datasets such as satellite imagery and the movement of seabourne freight to predict commodity prices for manufacturing companies.
I learnt a lot about politics in investment banks. The bonus culture creates incentives that make people compete with each other in a way that doesn’t necessarily benefit the organization. What I like about start-ups is that people are incentivised to co-operate both within and between companies. I’ve really enjoyed the feeling of everyone is in it together in all the early-stage companies I’ve worked at: instead of competing for bigger slices of the pie, we are collaborating to make the pie bigger.
A more positive learning experience from the world of banking was the professionalism and work ethic that my colleagues had. Things like communicating clearly or being punctual are pervasive in these places and I have occasionally found lacking in start-ups. At ChAI, I make it clear that both of these, and many other big, well-established firm’s policies are just as important in small teams.
Perception leakage is very important – if someone makes mistakes in a short email or turns up five minutes late for a meeting – you start to worry about other things they are meant to be doing (for example their ability to forecast commodity prices). So my view is that these are low-cost ways of signalling the company’s overall abilities.
When working in the ratified, well-resourced world of systematic hedge funds it struck me that less scalable organisations (e.g. manufacturing companies) didn’t have access to the datasets or skillsets that hedge funds do. Furthermore, the futures markets that we were speculating on had actually been set up to absorb risk from these manufacturing companies in the first place, but most of the activity (well over 90% in most markets) is speculative. Very little of it is companies that make things mitigating risk. So it seemed like a huge opportunity to democratise access to the tools the well-resourced hedge funds and bring them to those with physical risk (organisations involved somehow in production / manufacturing) with a service that, because of the economies of scale, would be cost-effective to sell as a third-party service.
The idea first came about when I participated in a trade mission with the UK government selling AI to China. This was a very humbling experience from a technical perspective, but did help me form a relationship with a Chinese state-owned enterprise that made pipes out of copper. This company bought copper at a variable price, sold it at a fixed price and absorbed that price risk without having any expertise on how to handle it. I built a prototype that forecasted the price of copper for them so that they could buy futures on the Shanghai futures exchange to offset their price risk when my forecasts were that the price would rise, and vice versa.
We are extending our services from intelligence (forecasting price movements with accompanying explanations) through decision making (advice on how much to buy and when) to an ability to actually offset the risk through us. In some ways we are selling a problem to people at the moment with our material price inflation estimates. We need to be able to provide the solution to this problem too through an insurance product that removes material price risk from physical supply chains.
We will be much more involved in the decarbonisation of supply chains. We already have the ability to advise corporates on how to alter the composition of their products in order to minimise their costs. We would like to extend this in a way that also considers their carbon footprint. Furthermore, we accelerate the development of green markets and adoption of environmentally friendly raw materials as users favour our more environmentally friendly commodities.
I have a very technical background and have always loved maths, computers etc. However, my real passion has always been to actually make things. I am much more an engineer than a scientist (as anyone who has ever seen any code I’ve written will confirm). I don’t see the point in theory without it resulting in something that actually impacts peoples lives. Furthermore, what I’m probably best at is linking between difficult theoretical ideas in unusual areas and problems that I know exist in the real world. And this skill doesn’t seem to depend on the domain that much (I’ve had a go at supply chain management, algo trading, portfolio construction, fine wine pricing and even diagnosing heart disease). A slightly less vainglorious explanation is also that I’ve never been the best at either the theoretical side of things or the practical, so being reasonably good at both is all that I am capable of.
I often feel that this is also true of being a CEO – I’m not the best at anything (other than storytelling) in the company, but I’m probably the best all rounder.
I am no longer involved in any other companies and actually haven’t been for 3 years. I used to take great pleasure in having a portfolio career and trying to do many things at once. I very deliberately have taken the opposite approach with ChAI and put all my eggs in one basket. This means I build up the frustration tolerance to deal with problems as they arise and not just flit to another project. More importantly, I feel a great deal of trust was placed in me by our investors and the incredibly high calibre and well-motivated team who have chosen to work for us. I would therefore feel like I was letting them down if my focus wasn’t fully on what we are building together.
More generally, I actually think that the search for a portfolio career is a mistake, and when people desire one it’s because they haven’t actually found what they really want to do. I am lucky that after almost 20 years, I have found what I really want to do: build ChAI.
The feeling that even conceptually we aren’t solving a problem worth solving or that people will pay to solve. I.e. that it’s not so much we haven’t found product /market fit, but that we never will.
I respect the ones that treat those who they work with, with respect and are bringing all the people in the company they have built along together, instead of treating themselves as the apex of a personality cult. I also dislike overly bombastic entrepreneurs because they create an arms race of panglossian exaggeration that makes it harder for those who treat truth with more respect to be heard against the noise. I think one of the implications of what is happening with Elizabeth Holmes / Theranos will be that the trade-off between being realistic and appearing ambitious will become balanced closer to where I am comfortable.
Lastly, I have a lot of respect for serial entrepreneurs who build successful businesses again and again. This shows that it’s not just luck on their side.
One’s stress response is arbitrary in terms of how appropriate it is for any given situation, so it always makes sense to try to reduce stress and not take anything too seriously. Nothing we do at ChAI will ever save someone’s life, so despite being important and valuable, it’s never worth worrying about things too much. Furthermore, being mildly optimistic is the most rational outlook one can have – one influences decisions by assuming positive outcomes and working towards them. This is really really important as an entrepreneur, you have to constantly take slight risks with the assumption that each gamble will have a positive outcome – and the accumulation of all these bets is what a company ends up being.
Another thing is to remember just how wrong I am about things frequently. It’s appropriate for CEO’s to be slightly overconfident and over-value their abilities. However, this needs to be tempered with a low-friction ability to change one’s mind based on others’ feedback. I am very fortunate to have a strong co-founding team to keep me on track here.