Ask anyone who regularly interviews people and they’ll tell you that few things are stranger than when the tables turn and you’re the one being interviewed. This is especially true when the person interviewing you has been dead for a year and a half. But here we are. Virtually Parkinson is a new podcast in which celebrities are interviewed by an AI model trained to speak and act like the late Michael Parkinson. The announcement of the podcast last year prompted a flurry of vaguely apocalyptic reactions. It was sacrilegious, some said, tantamount to digging up and reanimating a national treasure against his will. It was pointless, others said – of all the transformative ways to use AI, you’re blowing it on a podcast? Then there were folks like me, quietly nervous that Robot Parky was coming for our jobs.
On that last grumble, at least, I don’t need to worry yet. Primarily, this is because running Virtually Parkinson is a gargantuan operation. A normal interview traditionally only needs two participants. This, however, requires a small army. There’s the subject and interviewer, plus an engineer, a researcher and a clutch of producers holed up in a control room whose roles include making sure the AI sounds like Parkinson, that it asks the sort of questions he would ask, has an encyclopaedic knowledge of both the subject and Parkinson, doesn’t glitch, doesn’t repeat itself, doesn’t interrupt anyone in the middle of an answer and – most importantly – doesn’t overuse the word “fascinating”.
“It’s a full room but we do actually need everyone here,” says Mike Parkinson, the presenter’s son and one of the podcast’s producers, along with Ben Field and Jamie Anderson. “It’s too much work for one person. Ben’s got enough on his plate making sure everything’s working. He needs me and Jamie to be the eyes and ears. There’s this idea that we’re creating this thing that’s going to strip away human jobs, but we’re not. We’re probably using more people to work on this than we would if it was a talk show conducted by a human.”
Mike’s involvement is especially intriguing. Parkinson Sr only died in 2023, so you’d think it would be extremely odd for Mike to be around his dad’s voice every day, hearing him say things he never actually said. Not so, says Mike. “I’m a producer,” he explains. “I’ve worked – sorry, I worked – with my father for nearly 20 years. At work he became Michael Parkinson, but outside of that he was my father. So that’s not my father’s voice on the podcast. It’s Michael Parkinson’s voice.”
Do you think he would have approved of his voice being used this way? “My father was a very traditional man,” says Mike, “but when it came to his job, he surrounded himself with people who were much more on the cutting edge. My relationship with Ben and Jamie came out of a conversation I had with my father about what would happen to his archive when he was no longer with us. It’s a wonderful archive that he wanted to preserve – not because of his ego but because of the people he interviewed. And I said, ‘Well, look, there is AI.’ And he said, ‘Oh, that’s interesting.’ We batted it about but he was very open to it. It helps to keep his archive relevant.”
The ethics of it are at the forefront of the team’s thinking for other reasons. The 2023 Hollywood strikes were largely rooted in the fear of unscrupulous producers using AI to eliminate human writers and actors. Virtually Parkinson is an attempt to show that this technology can be used ethically. For example, all the interviews used to train AI Parky have been properly licensed. “For about 18 months,” Field says, “I helped various organisations across the media write policy on the legal, ethical and responsible use of AI. But last year, everything stopped being hypothetical and this podcast was commissioned. It’s a significant change.”
There is, of course, another reason why I don’t fear for my job yet. This is that, as an interviewer, AI Parky is no Parky. Although I’m told previous episodes have gone amazingly well (the comic and Strictly winner Chris McCausland apparently screamed “This is fucking brilliant!” after his interview) the one I witnessed being recorded had a much more muted response.
The subject was Monty Don, who entered the studio at the end of a long day promoting his new TV show Monty Don’s British Gardens. He seemed entirely unruffled by the weirdness of being addressed by the disembodied voice of a dead guy. But although he offered up a few interesting nuggets about his life, he appeared to grow more and more detached as the interview progressed.
“It was less satisfactory and less interesting than I thought it was going to be,” Don tells me after his interview. “There’s no response to what you’re saying at all. There’s just a pause and then another question. Look at the way we’re talking now. I just said something and you slightly raised your eyebrows. That’s important communication. Even if you’re doing radio, people will laugh or butt in or respond. So I’m interested in what they’re doing here, but there’s a long way to go.”
This doesn’t mean Virtually Parkinson is a dud. After Don left, the producers hopped behind a mic to deconstruct how the interview went from an AI perspective. After every interview, AI Parky is overhauled and improved. They noted Don’s claim that the interview seemed like various iterations of the same question, and mooted the idea of giving their Parky more of an agenda. In this sense, the series is an evolution. In theory, the Parky you hear at the end of the series will be much better than the one at the start.
“The whole point is to explore the relationship between AI and humans,” says Field. “You cannot make this show in any other way. This podcast will spark a conversation, and that’s what entertainment is supposed to do. You may disagree with the intent and think it’s exploitative to be using a national treasure’s voice this way. But I don’t have that opinion and Mike doesn’t have that opinion, though I fully respect that others do.”
Even more interestingly, Field mentioned a future in which interviews might be conducted with two simultaneous AI interviewers: a “conscious” asker of questions and a “subconscious” one that monitors the interview for mood and the passage of time, reacting accordingly.
Then it was my turn. My interview was less expansive than Don’s, largely due to the fact that, while AI Parky was fed 15 pages of in-depth research about the great gardener, all it had on me was two paragraphs from my publisher’s website. Nevertheless, it was enough for me to experience the utter weirdness of Virtually Parkinson.
“Good evening, ladies and gentlemen,” the instantly recognisable voice began. “My guest this evening is someone who has carved out a unique niche in the world of journalism and creative writing. Please welcome Stuart Heritage!”
I’d put a lot of money on the real Parkinson never knowing who I was, so to hear that voice introduce me was startling, almost too strange to overcome. Parkinson still feels part of the culture – and it’s immediately clear why he, more than anyone else, was chosen for this. The warm timbre of his voice is so human: I could even hear him taking breaths between words, which automatically offset the fact I was actually being coldly interrogated by a computer.
Parky’s first question was simple enough. Essentially: “How did you become a writer?” But it threw me. “Oh,” I said, like a rabbit caught in the headlights. “Um, I, er.” Eventually I found my footing, but it was hard to deal with the unbending format of AI Parky’s responses. You say something, he acknowledges the answer, summarises it, then asks another. “It’s interesting how stepping outside of familiar surroundings can offer clarity and perspective,” he said after I mentioned briefly living in South Korea. Such validating can be an important part of human conversation, but to do it so rigidly, regardless of what was said, can seem wildly insincere. It made Parkinson sound like a wouldbe “pickup artist” who’d read too many Neil Strauss seduction books.
There are other niggles. This may just be my own needy problem but, whenever I’m interviewed, I want the interviewer to like me. Like Monty Don, I was desperate for something – a nod, a glimmer, a murmur of agreement – to show that I was connecting. I suspect many of AI Parky’s interviewees will feel the same way.
At one point, probably down to the scarcity of information the model was given, Parky latched on to an award I won 20 years ago. “Tell me about it,” he said. “It’s a rectangular lump of metal,” I replied. His response, basically that rectangular lumps of metal are fascinating, made me laugh out loud.
Weirdly, the most human thing that happened came right at the end. As the interview concluded, AI Parky thanked me and I thanked him back – then something glitched. “YES STUART?” he suddenly barked, completely out of nowhere. Which was nothing really, just a fragment of code gone awry. But it felt organic and informal in a way that nothing previously had.
Plus, it was very clear I was not talking to the Michael Parkinson. To watch classic Parkinson in his pomp was to watch a master at work. He knew when to stick to the prepared questions and when to chase his interviewee. Watch any of his interviews and you’ll see celebrities scurrying to win his approval. There’s none of this with his robot counterpart. Crack a joke, attempt a diversion, refer to the strangeness of the situation, and the model will just blankly move on to the next interaction. It’s genuinely unsettling.
But, as Field points out, the technology is evolving at a remarkable pace. When I ask if we might one day have a visual representation of Parky for subjects to react to, Field replies: “We’re a bit off that.” Is it years away, decades even? “About five or six weeks,” he says. Like it or not, the future is already here.
• Virtually Parkinson is available on YouTube and via podcast providers from today
