top of page
  • Writer's pictureMadeleina Kay

"Good luck with your cursed life" - AI and YOU Podcast Released

Updated: May 2

The podcast is finally out! (Well, the first episode, at least…) It’s been an incredible experience hosting a podcast for the first time and I’ve learnt a lot during the process...

 When I was initially contacted by the producer (who I had never met before) with the offer of this gig, hosting the English language version of a multi-lingual podcast on the topic of AI – my gut reaction was ‘No way! I know fuck all about AI… Who on earth suggested me for this role?!’ (I later found out it was a Brussels-based journalist from Vox Europe - who I have known for many years).


I responded in a more diplomatic way to express my reservations that I wouldn’t be comfortable hosting a conversation about a topic on which I wasn’t an expert. I was informed that this was no problem whatsoever, as the 10-episode podcast series was an entirely “pre-scripted monologue”, meaning that no prior knowledge of the topic was necessary. So, I tentatively agreed, since it was paid work and I was keen to fill my void of knowledge on the topic of AI. What I didn’t realise was that I was also going to learn a lot about “voice work”.

We had planned to record all ten episodes during a trip to Brussels in early March, however, due to delays in the script writing – only the first five scripts were ready to be recorded by then. I had contributed heavily to the editorial work, since the scripts were not written by a native English speaker, but also content-wise, the scripts were personalised with anecdotes (for example, I mentioned my high school history and geography teacher in the episode on education).


Since the company, Europod, was Brussels-based, I assumed they had staff and a recording studio in Brussels – hence making the initial suggestion to record there - but, I was wrong (apparently they used to before going fully remote). I discovered, whilst on the Eurostar to Brussels, that the studio which they had booked was a “DIY” studio out of town (meaning the public transport was so problematic, I was told I would have to take an Uber to get there). I immediately panicked – I have no sound tech knowledge. When I record my songs, I have always worked with a producer. I was concerned that if I attempted it by myself, I would totally fuck it up.


So, at very short notice, the producer found a freelance podcaster to support the recording and act as a vocal consultant. She was lovely, and really helped me to understand how to deliver the content. Focussing on the tone, pitch, emotion, stresses and pauses - in the same was as I would when recording a song. The only advice she gave me which turned out to be unhelpful – was to speak very slowly (which isn’t an easy thing to do, as it is unnatural). Much to my exasperation, the producer told me afterwards that I needed to pick up the pace…


It wasn’t clear before we arrived, how the “DIY” studio worked. Unfortunately, we didn’t realise we needed to bring our laptops – but it turned out the recording devices could record directly to an SD card – so we thought we were saved. No such luck. After recording for six hours, the vocal consultant returned home to discover the files had corrupted and we lost all the recordings. After contacting the studio, she discovered that their devices only work with a specific type of SD card (which they sell). This wasn’t communicated anywhere, so how were we supposed to know?


She told me afterwards, that this incident brought her to tears. I was so exhausted by that point that I was in “meh” mode, shrug-and-whatever, I guess we have to do it again – at least nobody died. Worse shit has happened.

So, the day I was due to leave Brussels, we travelled to the out-of-town recording studio once again. My travel curse seemed to have bugged my Uber app that day, as it took fifteen minutes for an Uber to arrive and when it finally appeared at the end of the street, it stopped and a couple with a baby in a pram proceeded to get in. I checked the licence plate – it was definitely our Uber. After a few minutes, the driver realised the mistake and the couple got out and unloaded their pram – it was all rather strange. On the way back (after re-recording all five episodes directly to a computer this time) the Uber situation was even worse! I cancelled the first Uber after it didn’t arrive for fifteen minutes. The second Uber sent us a message after five minutes to say the previous customer had vomited on the seat and they had to clean their car. The third Uber stopped at a gas station for ten minutes, so I cancelled and called another only to be connected to the same driver. He showed up eventually, but we had been waiting forty minutes by this point (fortunately I had ample time to catch the Eurostar). The vocal consultant was baffled ‘I’ve never had these problems with Uber in Brussels’ she informed me. I told her I had a curse and gave some examples of my recent travel disasters. Her parting words to me were ‘good luck with your cursed life.’

The remaining five episodes were recorded in Sheffield, after the editing process had been finalised (or so I thought). We booked my producer at Big Dog Studios, who produces all of my music, since he has a lot of experience producing podcasts / audiobooks for Audible and BBC too. Had I known the company did not have their own studio in Brussels, it would have made so much more sense to record all the episodes in Sheffield… But anyway, hindsight is a wonderful thing. This process went much more smoothly without any of the SD card or Uber app nonsense – it seems kind of ironic that there were so many technical hiccups when we recording a podcast on the topic of AI…


Due to revisions to the scripts made by the media partners, I had to return to Big Dog Studios a final time, to re-record a few lines and to record a trailer for the series. Overall, I think the content of the series is extremely interesting and informative, covering a huge range of topics, including; jobs, education, healthcare, climate change, war and security. I was very happy to be part of the project and I genuinely hope listeners enjoy it as much as I did.

When I visited Rome, I met the producer who lives there – he came to watch my talk / performance event and we had a long chat about the podcast afterwards. He gave me his ear buds and told me to listen whilst he played the (then unreleased) podcast. After a few seconds of what felt like torture, I took the ear buds out and told him I couldn’t stand the sound of my own voice. He replied, "but you're a musician" and asked me if I thought the production was bad? And I told him that it was great, I just hated listening to myself – probably because I get trolled all the time about my voice ("drowning cat", "howling wolf" "screeching mother-in-law" etc. etc.). His response was ‘And you let it affect you?” – a question to which I don’t have an answer, I guess it was rhetorical. I laughed it off and said I would listen to the French version of the podcast or possibly the German, although I wouldn’t understand much of it.

On a related note, I recently had a conversation with Ukrainian woman (currently living in Poland) who informed me that she is "obsessed with English accents" - apparently because she watches a lot of period dramas. I asked her what she though of my accent (as most people are unable to place it regionally) and she described it as "clean". I pressed her on what she meant by this and after some discusion, she said, "posh, but sophisticated not braying like the politicians" (I assume she meant Boris Johnson and Jacob Rees-Mogg). I guess I'll take that as a compliment.


The only job left is the marketing promotion of the series. There is a weekly episode release – So, expect some A related social media spam in the coming weeks.


I had the idea to ask AI to generate an image of me hosting a podcast - to see what it would come up with... There are a rather unhealthy amount of photos of me on the internet (although I have no idea how the data set training works), so I wanted to see if it could generate anything which looked vaguely like me - given that when I asked Leonardo.AI to generate an image of Rishi Sunak, it bared little resemblance - my expectations were low.

Prompt: 'Madeleina Kay hosting podcast'

Prompt: 'Madeleina Kay hosting podcast blue hair'

Prompt: 'Madeleina Kay hosting podcast blonde and blue hair'

 Prompt: 'EUsupergirl hosting podcast'

Prompt: 'EUsupergirl hosting podcast blue hair'

I think the first of these five images looks most like me, in terms of face shape - you'll never see me dressed like that though - I have one black jumper and it is covered in beads. Number three's hair is most similar to mine, although I would like to have number five's hair - my hair never grows past shoulder length (I haven't had it cut for two years and it has stayed the same length) I think because all the peroxide I have used causes it to break.

I thought it was funny how three of these images have got wonky eyes - because I have a congenital condition 'Duane's Retraction Syndrome' which means that a muscle is missing, so it often looks like one of my eyes is on the piss. To be fair, most people don't notice since I usually compensate by turning my head, as it's slightly painful trying to look in a direction my eye won't turn. But trolls, of course, have noticed it and mocked me for my "skelly eyes" before, whilst speculating about whether or not they would rape me - "if she kept her eyes closed". Anyhow, I guess it's just an AI glitch but it made me laugh that if not anything else, they got my wonky eyes just right.

Still from Betrayal In Your Eyes Music Video.

2 views0 comments


bottom of page