The startup accelerator model is as vulnerable as many other fields to being shaken up by AI.

And if it’s an accelerator in the higher education spinout world, it’s also exposed to the inescapably shaky finances of UK universities today.

To explore these topics, I spoke to Ben Clark, director of the spinout-focused Future Worlds accelerator, which helps research-based founders from the University of Southampton and nationwide.

Read on for our conversation. But first:

• Shout out to Voxelo.ai, which we covered 13 months ago, right at the start of their journey. They’ve now closed their pre-seed round at £650,000. Find out more about their real-object-to-3D-graphic tech on their website.

• Premium subscribers to PreSeed Now can explore updates like this about the hundreds of startups we’ve profiled in our Startup Tracker.

– Martin

What should accelerators be in a world where AI can mentor startup founders?

Ben Clark has been in the entrepreneurship world for a long time. He eventually gravitated towards marketing roles in tech and then to the Future Worlds accelerator at the University of Southampton.

Starting as a mentor in 2017, he eventually became director of the programme in 2019.

Future Worlds is principally focused on supporting Southampton-based researchers and students to launch and scale startups and spinouts. It also runs nationwide programmes for AI and silicon photonics spinouts.

Aside from this, Clark fits in being an Associate Dean at the University of Southampton, and an AI podcast host.

I spoke to him about running an accelerator for spinouts, how AI is transforming the role of accelerators, and how investment in deep tech masks crumbling foundations in the university sector.

This conversation has been edited for clarity.

MB = Martin SFP Bryant, BC: Ben Clark

MB: What are some of the most notable startups we should know about from your time at Future Worlds?

BC: There’s always the risk of picking your favourite child. We've helped hundreds, into the thousands, of people exploring projects. For us, a key reason for existing is that there's a load of latent potential across the university student and research base, particularly at PhD level and above. There’s a real sweet spot that has potential but wouldn't otherwise happen. It wouldn't get into the real world if it wasn't for some kind of catalytic intervention.

A good example would be Aquark Technologies. It came from a PhD in physics. The founder developed some cold atom technology. He turned up saying ‘I've got this invention. Do you think anything could happen with it? I’ve no idea where to turn.’

They spun out in 2021, and about 18 months ago they raised a €5 million seed round from the NATO Innovation Fund. They have gotten incredible sea trials with the Navy, they've done trials with the army, flown this technology on a drone to demonstrate capability.

This is about bringing quantum sensing and timing technology to the market in a way that lots of people have been able to hypothesise that this should be possible. Lots of people have been able to prove capability in a lab, but what they've been able to do is miniaturise it, strip the cost out, make it rugged, and make it actually work in the real world.

This was a Romanian who is brilliant at physics, gets a scholarship to come and study undergraduate physics here, then an opportunity to do a PhD here.

He has no network or background that has any relevance to doing startup stuff. He can't do a friends and family round where they're going to drop a few million to help them get started. But through going through the route that we have taken him through, and tapping him into our network, we've been able to then help him go on this completely transformational journey.

Another would be Viridi. This came out from chemistry, and has developed a catalyst that can be used to convert waste CO2 into a feedstock for plastics and a whole range of other FMCG goods. It’s now in proof-of-concept trials to take it out to an industrial scale with a number of big chemical companies.

Another would be AccelerComm, working in the 6G space. Their technology has been a core underpinning of the first 5G base station on a satellite, providing 5G mobile coverage, mostly from a low-Earth orbit satellite.

MB: AI is transforming early-stage startup creation. How is it changing your role as an accelerator?

BC: I was with a startup on Thursday, led by a serial founder. He's been in stealth mode for the last six months, having exited his previous venture. He was saying ‘look what we've been able to build in six months leveraging things like Claude Cowork and Claude Code. It surpasses what we built in over four years in the previous company. We are just moving to products so much more quickly, and scaling much more quickly’.

We're seeing academic teams that are able to automate a whole lot of their processes, and their customer discovery work, a load of their market analysis work. They're leaning into even just some of the generic understanding of the process of building a company, like what investor readiness looks like. They can access that kind of information far more easily.

From an accelerator point of view, we're certainly at that point of asking what our secret sauce is. What really adds value to founders that isn't automatable, that you can’t just tap it into ChatGPT or another large language model and get the same information? And where should we just lean into the automated tools giving the founders access to knowledge?

If I'm honest, that's still an open question.

There's a whole load of stuff that we are no longer doing, because they can get it through AI, but then there's a whole lot of stuff that if you just leave it to AI, it gets you maybe 80% of the way there. It gets you generic information.

So much of what AI doesn't do is often the localising information for the individual context, whether it's UK university stuff, or specific dynamics around the investors that you have access to, or how you actually connect with those investors.

Investors are deluged by startups that can get an automated, good looking pitch deck. But the reality is, there's often no substance behind it. That means building genuinely transformational relationships with investors actually becomes harder, rather than easier, because they've just got noise.

And so a lot of that stuff is where there is still a really important space for us. Most of our founders are first-time founders. We're badged as an accelerator, but it's very much early stage incubation that we're doing, and a lot of that is helping people with the fear of the unknown, helping people with the confidence that they can navigate this, stopping them feeling like an imposter.

It's giving them the confidence that, yes, you can do this, we can hold your hand through this side. Some founders can get a lot of that from AI in terms of the practicalities, but a lot of the softer stuff and the personal stuff is where the accelerator can really help to add value.

MB: What are some of the things you’re not doing anymore as a result of AI?

BC: That would be just some of the basic kind of workshops on basic knowledge, like how do you do banking? Where should you look for legal advice? All that kind of stuff.

It is so easy to get a load of that more basic information. So we’re encouraging people to lean into that more strongly, rather than us doing what can otherwise be quite time intensive work around some of that stuff.

You can invest a lot of time in doing bespoke content for people, and a lot of what we're seeing is that we probably need to do less of that, because they can get 90% through Claude, and then we'll just do the extra bit around it.

MB: It feels like there’s more support than ever for deep tech university spinouts these days. But what are your thoughts about the spinout landscape in the UK right now?

BC: I think you're right. There is a lot of support that is available. I see it as a remarkably positive thing to have more support for more people.

Yes, at a high level, there's a lot that is going on and that's great. But at a low level, within institutions, the finances for universities across the UK are really, really tough at the moment. The student fee income often hits the headlines. The impact on international students coming in can have a really big impact on university finances.

The whole university system is creaking financially and that is a real challenge for academics and researchers to have enough latitude and capacity to then indulge in this exciting startup space.

We're embedded firmly within deeply technical faculties and schools where maybe they've got shedloads of students that they're having to deal with, because they just need to make the numbers work. So they need to bring in more students, and that means the researchers have less time to then spend on some of that early-stage development, and so that makes it really difficult.

There's still a lot of appetite to do that, but often the practical constraints can make it more difficult, and I see that getting more difficult over the next couple of years because of the state of the finances of universities.

On the upside, it is pushing more universities to think more broadly and more laterally and say ‘if we can really help to launch some great spinouts and some great startups and really commercialise this wonderful IP well that becomes a valuable income stream’.

That can be a good thing, but that can also lead to rent-seeking behaviour on behalf of universities, where they're trying to extract more value from these early stage startups than really makes sense to support them as an early-stage startup.

So it's like the classic thing. You need angel investors who are investing because they can afford to lose all this money. You don't need angels who are investing expecting a deep tech startup to be revenue generating within 18 months and delivering them a lovely exit within two years.

Perversely, universities can sometimes act like that angel investor that they're expecting a return far too soon. So there's still work to be done in that area, but I think things like the USIT Guide and the government's independent review of spinouts has helped to encourage universities to be more outward looking and more investor-friendly and founder-friendly. All of those are good things.

MB: We’ve seen the government pledge to support spinouts, and quantum technology in particularly recently. So it feels like a lot of good stuff is being built. Unfortunately, it’s on top of the crumbling foundations of the university sector.

BC: Don't get me wrong. I have advocated very strongly, internally and externally, for universities being more founder- and investor-friendly and being more liberal in their approach to just releasing the good stuff.

I did a Times Higher Education article when the spinout review was going on, saying that when a deep tech startup from a university spins out, it is unproven, often not much more than an idea in a lab. Don't be haggling over the spoils, because there are no spoils at that point. It's at the start of a journey that's going to take blood, sweat and tears to maybe get somewhere at some point in the future. Just do all you can to help them to succeed.

I also sit now in a leadership position within a large university, and I do see the reality on the ground, and it's easy for politicians, for investors, for externals, to wade in and kind of criticise universities for not being more enabling, and often they're completely unaware of the financial reality. Deans or vice-chancellors are having to balance the books, and it is intensely difficult at the moment.

So yes, fixing the foundations is something that it is in the government's gift to do. There should be more funding put into this space centrally to enable that.

The government are throwing billions at research through UKRI. What they need to do is be strategic around how they use that funding, to ensure that the foundations are secured, but also then that universities are incentivised to get behind this stuff, because often that's where there's a perversity in the incentives. Yes, there's funding, but the funding may go to the wrong places and not line up to incentivise the individuals and the institution.

For example, funding for spinouts to buy access to facilities in the universities in order to build an early proof-of-concept and to do some of the highest risk R&D is a thing that can be funded through some of these mechanisms. That can make it more attractive to universities to make those facilities available, while de-risking the university's finances and ensuring that IP goes from the university into those early-stage deep tech startups that will hopefully be the champions of tomorrow.

MB: Finally, what’s next for Future Worlds?

BC: We had the AI Demo Day recently. In June, we've got a demo day for our general cohort that is across technologies and sectors, and includes silicon photonics.

At the end of April, we're taking our cohort out to San Francisco and the wider Silicon Valley. We've got one from the AI cohort joining us there, and a number from Southampton and several from Silicon Photonics.

Most of them have never been there before, so we’re opening their eyes to the massively different mindset and culture that you experience out there, lifting their eyes to see just how big things can be.

Often that will be about them bringing that mindset back and then launching and scaling in the UK. Sometimes that is about connecting them with an accelerator out in the US so they can launch and scale out there.

How are the startups we’ve previously featured getting on? As a Premium subscriber, you get access to the PreSeed Now Startup Tracker to find out: