A thesis that nobody's read, narrated by someone who isn't there, for an audience that doesn't exist.

An AI programme has apparently made a podcast out of my PhD thesis, in what seems to be a Guinness World Record attempt for 'most pointless endeavour ever'.

I’ve done a lot of podcasts in my time. Because of course I have; I’m a middle aged straight white guy in 2025. “Would you like to be on my podcast?” is how my people say “Hello”.

I’ve guested on more than I could ever hope to remember, often repeatedly. I’ve hosted a fair few as well, including Brain Yapping (with Rachel England) and Why Does This Thing Exist (with oldest friend Simon Feeley). I’m also co-host of the Patreon-exclusive SMERSH spinoff The Monday Night Movie Club (with John Rain and Dan Thomas).

My podcast output has been rather eclectic, to say the least. From mental health musings to deep-cut analysis of decades-old anti-smoking ads, from guidance for parents about technology to wildly explicit accusations about the personal proclivities of celebrated botanist David Bellamy. I like to think I’m quite flexible with the subjects I’m willing to discuss for the entertainment of others.

But as diverse as they may be, there’s one thing all my podcast contributions have in common; I was actually there! As a willing participant in the proceedings.

You wouldn’t think I’d need to clarify that. I wouldn’t have thought it necessary, until very recently. But I do. Because it is.

Because I recently got this email from the site Acedmia.edu.

I know I posted this already at the top of the article, but here it is again anyway.

If you don’t know what Academia.edu is… honestly, neither do I really. From what I can tell based on a cursory web search, it’s ostensibly a platform for researchers around the world to upload and share their research freely, away from the stranglehold of the official academic publishers. Which are, in fairness, an absolute racket.

In practice, it seems it’s more like the answer to the question “What if LinkedIn was inhabited entirely by academics?” Is that unfair? Possibly. Possibly not.

In my defence, given that I’m a science person with a moderate media profile, I get a LOT of emails. Often from people and organisations I don’t remember ever engaging with. Probably because… I haven’t? Even so, I get so many press releases and announcement and suspiciously vague conference invites¹, that I just roll my eyes and click delete. The barrage of emails I got from Academia.edu presumably just got mentally lumped in with all the other pseudo-spam.

Which is probably good. Because Academia.edu email you a LOT. Anything that even hints at involving you in some way, they email you. Someone looks at a paper of yours? Email. Someone clicks on a link that includes you? Email. Someone with a similar name to you publishes something? Email.

That last one I’m not too annoyed about, because it led to me learning that there’s another Dean Burnett out there, based in South America I think, who does research into ‘laser dentistry’. He is clearly a much cooler Dean Burnett and you should go and read his blog instead.

…still here? Ok, fair enough. Takes all sorts, as my gran would say.

But yeah, my blasé attitude towards Academia.edu meant I was blissfully unaware of the more… ‘shady’ aspects of the site. The dubious use of the ‘.edu’ domain. The questionable business model. The ‘you can pay for more attention to your research’ efforts. All that passed me by.

But then they emailed me, to tell me that they’d had an AI programme make a podcast out of my PhD thesis². And that struck me as wrong, for so many reasons.

I’ve often joked that my PhD thesis is technically the first ‘book’ I ever wrote, but the one that has been read the least. Because the people I know have read it number in the single figures. They include me, my two supervisors, my two external examiners, and one very nice Argentinian medical student who emailed me a few weeks ago asking for it as he’s a fan of mine and wanted to read my ‘actual’ research.

My aforementioned and beloved late grandmother Brenda also insisted on having a copy when I graduated as a doctor, and insisted “I WILL read it!” despite my repeated reassurance that she didn’t have to. Because it wasn’t a book. It was the most jargon-heavy description of original research, which she had no experience of whatsoever, that I was capable of producing. But she was determined.

Four weeks later: “I’m sorry Dean, I couldn’t read it. It didn’t make sense!” Bless.

But it turns out, if this had happened now, she could have just downloaded a podcast summary of my PhD! Spoken by an AI programme. What a boon for everyone!

Except, it’s not is it. Because, truth be told, I have questions!

Photo by Seej Nguyen

First and foremost, WHY would anyone want this?

Who is the audience out there who are so interested in my output that they want to know more about my nearly-two-decades-old PhD research into an incredibly specific and technical area of neuroscientific research, but simultaneously not interested enough to find and read it, so prefer have it narrated to them by a creepy synthetic facsimilia of a podcast host?

Secondly, what is the point of this endeavour at all? Because it’s not just me. Any ego-fuelled suspicions I had about someone trying to profit from my meagre name recognition were quickly dashed when I posted about this online, and numerous fellow doctors, in all manner of fields, replied to say they’d received similar emails. It’s a global-community-wide thing.

But, again, why? Where is this mass audience for soulless audio summaries of the niche-st of niche research? Maybe I’m naïve and times have changed, but I’d assume that anyone who’s at the point where they need to explore a random someone’s PhD thesis is already deeply invested in the subject matter to need all the minutiae and details, something bound to be absent from and AI podcast summary. And that’s assuming what is included would be accurate. Which is far from guaranteed.

Thirdly, why the hell is a podcast about my thesis, which took five years of my life to produce, less than five minutes in length? If it’s a brief summary, like the actual abstract, it should be over in about 30 seconds. If it’s a full verbal reading, it should take a few hours. But 5 minutes? That suggests… more AI ‘adlibbing’ than I would ever be comfortable with.

And fourthly, if you’ve taken my work, something that, again, took me five years to produce (I wasn’t the best researcher, by any metric), that I don’t think I ever uploaded to your site, without my permission, without my approval, without letting me know about it, without compensating me for it, and using it to create an AI podcast, something which I would and do not approve of… why would you then email me about it as if you’ve done me a favour?

Someone who saw this did tell me that the email was letting me know the podcast had been created, and I’d need to log in and approve to have it released. Which is fair enough. But that still begs the question, why?

What is going on? Why would anyone take a dissertation (or multiple) that nobody has read, and use it to create a podcast that nobody will listen to, narrated by a voice that doesn’t really exist?

I realise I could answer a lot of questions here by actually listening to the podcast, but with all due respect, fu*k that! And f*ck you too if you honestly thought I’d be willing to even tacitly endorse this baffling and frustrating endeavour.

I know there are many tech-minded people saying we must embrace AI, because it’s inevitable and will redefine everything blah blah blah.

Maybe they’re right, and that’s all true. But that doesn’t mean I have to approve of this! And if, even at this early stage, AI is being used to create podcasts on published works, for academics who didn’t ask for them, to benefit audiences who aren’t there, I’d say that doesn’t bode well for it’s supposedly glorious future.

I would never suspect that Academia.edu are behaving like every other company in 2025 who. They say ‘When all you have is a hammer, everything looks like a nail’. Well, every modern organisation seems to think “When you have an expensive AI, everything that anyone does is a thing that must have AI applied to it”. That’s not as catchy, though.

Anyway, that’s a rant. Now I’m going to go and cancel my Academia.edu account.

---

Please buy my latest book Why Your Parents Are Hung-Up on Your Phone and What To Do About It, before they turn that into a pointless AI podcast too.

1. I’ve attended a number of academic conferences over the years of all stripes, but I’ve never encountered an official one where the organisers invite you by referring to you as your email address. As in, “Dear wedontknowyourname@justchancingourarmhere.mail.com. We would formally like to invite you to present your work…”. It seems to be a specific sort of phishing scam aimed at academics, and those who aspire to be one.

2. Or, more accurately, that ‘an’ AI had made a podcast of my thesis. Of it’s own volition, I guess? Nobody to blame, really