Towards the end of our conversation, Tony Raven brought up a short story.

E. M. Forster published The Machine Stops in 1909. Its people live alone in identical underground rooms. They speak through glowing plates, attend lectures remotely, summon music and literature through buttons, and depend upon a vast technological system that almost nobody understands.

One character wants to speak to his mother in person.

“The Machine is much,” he tells her, “but it is not everything.”

The story arrived at the end of the interview, but it could have sat at the beginning.

Raven has spent much of his career helping knowledge move between worlds: from physics laboratories into companies, from universities into public use, and from an idea held by a researcher into something that can attract investment and survive outside the institution where it began.

But he kept returning to what does not move so neatly.

The failed experiments nobody recorded. The reasons behind a decision. The conversation in a corridor. The researcher who knows why something works but has never written it down. The judgement that depends upon understanding another person’s motives.

Forster’s terror is not simply that a machine becomes powerful. It is that people become dependent upon a system they no longer understand. Life is convenient, mediated and organised at a distance. Knowledge arrives through the system. Relationships pass through the system. The body becomes almost an inconvenience. Judgement thins out. Then the Machine begins to fail.

The point now is not that Forster predicted screens, remote meetings or artificial intelligence. That reading is too easy, and too pleased with itself. The more serious question is what happens when a culture forgets how to meet, judge, repair, verify, preserve or act without the systems that have made life efficient.

That question sits oddly well beside a conversation with Raven. He is not a novelist, archivist or literary theorist. He is a physicist, entrepreneur and former Chief Executive of Cambridge Enterprise, the University of Cambridge’s research-commercialisation arm.

His career has the shape of a series of accidents that became institutions. He wanted, as a young man, to be a pilot. A visit to the Science Museum and an Ishihara colour test put an end to that. Physics led to Manchester, Oxford and laser-fusion research at Rutherford Appleton Laboratory; then Japan, consultancy, companies, Southampton and Cambridge.

By the time Raven arrived at Cambridge Enterprise, he had already been on both sides of the problem: the person trying to turn an idea into a company, and the person helping a university decide which ideas deserved support.

What comes through in conversation is an impatience with lazy myths about innovation. Raven does not describe invention as a miracle. He describes routes.

Who backs the idea? Who protects the founder? Who knows what the research was trying to do before it became a pitch deck? Who remembers the mistakes made before the result looked obvious? Who can pick up the telephone?

Raven’s sentence argues against a culture that thinks preservation means storage.

A paper can be filed. A patent can be registered. A PDF can be uploaded. A model can ingest text. But the living knowledge around an idea often sits elsewhere: in the researcher’s memory, in the laboratory’s habits, in the failed experiments, in informal warnings, in the reasons something was built one way rather than another.

When research leaves a university and enters a company, the danger is not only that the intellectual property will be mishandled. Context may be stripped away from the thing itself.

“If you just pick something up and try and run with it,” Raven says, “you can spend a lot of time learning through all the mistakes that the researchers have made to get to where they are.”

Cambridge becomes interesting here beyond the usual language of innovation “ecosystems”. The word is overused, but Raven’s version of it is practical.

A university is a dense social field: researchers, founders, investors, administrators, students, former students, mentors, sceptics, funders and people who know which door to knock on.

The route from research to impact is not a clean diagram. It is a series of conversations that preserves enough trust for risk to become possible.

Raven puts it bluntly: “Technology transfer is a contact sport.”

He is not romanticising proximity. He is describing a mechanism. A formal meeting requires an agenda and a time slot. A head around a laboratory door may produce the conversation that moves an idea forward. “There is much more social science in there than process,” he says.

That is a useful corrective to the way culture talks about technology. We talk about platforms, funding rounds, patents, datasets and models. Raven keeps returning to motives.

Why is somebody doing this? What do they want? Are they chasing money, power, attention or the right outcome? Who can be trusted to carry an idea without deforming it?

The same question applies to archives and publishing. What survives is not always the best work. It is the work that finds systems capable of carrying it.

A manuscript survives because somebody keeps it. A journal matters because somebody edits it. A public record remains public because an institution accepts responsibility for it.

On the surface, Cambridge Enterprise and Litro’s Future Archives belong to different worlds. One commercialises research; the other publishes literature and cultural thought.

But both are concerned with transmission. Both ask what should be carried forward, by whom, under what conditions, and with what evidence of origin.

From laboratory to institution

Eric Akoto: When you look back at how research moves from universities into the wider world, what has changed most?

Tony Raven: The big difference is the encouragement to do it. The British government gave universities permission to do this and supported it financially to get it going.

It has gone from being something people did outside the system to something adopted from the top. If you look at universities today, impact and interaction with the environment around them are central to their ethos.

Building ambition at Cambridge

Eric Akoto: What did Cambridge need when you arrived?

Tony Raven: When I started, we were putting about £50,000 into two companies a year. I said that if we were going to do this, we should do it with ambition worthy of a university such as Cambridge — the ambition MIT or Stanford would have.

We had to be in control of our own destiny. If we thought an idea was great, we had to be able to pursue it under our own control.

A lot of successful founders told me: I built a very successful company, it was a lot of hard work, people made a lot of money out of me, and I did not.

That was often because of the investment terms. So the structures we built were also about looking after founders and giving the company patient support.

What gets lost in translation

Eric Akoto: When research becomes a company, what tends to disappear?

Tony Raven: The important thing is the people. The tacit knowledge is probably more important than the codified knowledge.

If you just pick something up and try to run with it, you can spend a lot of time learning through all the mistakes the researchers have already made.

So it is important to maintain the relationship and maintain access to that knowledge.

Why social links matter

Eric Akoto: You said social links may matter more than formal technical links. What do you mean?

Tony Raven: A university is a mass of people researching all sorts of things — some complementary, some competitive, some directly applicable.

It is the ability to have conversations, to tap into the tacit knowledge of the individuals who created the technology, and also the community around them.

The Americans have a saying that technology transfer is a contact sport. A lot of people see it as a process. I always say there is much more social science in there than process.

It is about people, their motivations and the way you interact with them.

The supposedly closed university

Eric Akoto: What do founders misunderstand about working with universities?

Tony Raven: People say universities are difficult. I have never quite understood that. When I was running companies, I had zero problem connecting with universities, finding the best researchers and working with them.

The research is published. It is a very open community. Researchers go to conferences. You can talk to people and pretty quickly find your way to the person considered the world expert. You just need to put in the legwork.

The digital dark age

Eric Akoto: For Future Archives, I am interested in knowledge that exists but is not publicly searchable. What happens to that knowledge?

Tony Raven: Vint Cerf has talked about the digital dark age. We have medieval manuscripts written by monks sitting in libraries. But try to find something on the internet from the year 2000.

The website may no longer exist. If it was stored on media, you may no longer have a drive to read it. Google cannot search it. AI may not have it in its training data.

There is a whole body of knowledge that exists, or existed, and we may not be able to find it.

Where memory resides

Eric Akoto: Is there an answer?

Tony Raven: It has to be stored in some tangible, readable form. But there is so much of it. What do you need to store and what do you not need to store?

I have digital photographs going back years, and I have 35mm slides. The slides are fine. They can sit in a box and I can pull them out. I have a family archive of photographs going back to the middle of the nineteenth century.

But with digital material, what do I do with it? Will the storage company still be there? Will my relatives know it is there? Will they continue to pay the subscription?

Where is that memory going to reside in the long term?

Future Archives is not nostalgia for paper. It is a question about dependency.

The cloud feels weightless until an account lapses. A website feels public until it disappears. A search engine feels comprehensive until you look for something you know happened and cannot find it. AI feels authoritative until it invents the missing bridge.

The archive is becoming searchable before it is becoming trustworthy. AI can summarise, retrieve, imitate and recombine. It can also miss what was never digitised, misread what was poorly described, flatten provenance, or present an answer without a visible route back to human judgement.

Can AI be trusted?

Eric Akoto: Can AI be trusted as a knowledge system?

Tony Raven: I think it is slowly dawning on people that you cannot take AI at face value. You have to invest time and money checking whether what it told you is correct.

I am sceptical. I think it will have a long-term role. It is useful for encompassing statistics across databases beyond human comprehension. In medicine, for example, a doctor can only know so many diseases; AI can hold a much wider field.

But a colleague of mine calls it not AI but EI: emulated intelligence. It gives the appearance of intelligence, and we anthropomorphise it. But there is no intelligence there. There are no morals or ethics.

A name attached to the work

Eric Akoto: What is trust now?

Tony Raven: What is trust and provenance? It is that someone you trust has put their name to it.

There are news organisations and journals we trust because people we trust have done the research. That is their job: to do the legwork and tell us what is happening.

I am moving more and more away from social media, particularly for news, and towards subscriptions to organisations I trust to have done that work.

Who guards the record?

Eric Akoto: Who guards the systems that claim to guard the record?

Tony Raven: That is the question. Who guards your system to make sure it is doing what it is supposed to do, in the same way that you are guarding other systems?

You have to cast your net wide and filter for the people you want. A lot of people, particularly on the internet, are after money, power or influence.

But there are people who want to do the right thing. Those are the ones you need to look to.