On 29 July, Elliott Starr and Beatriz Zambrano announced the launch of ‘Alan Watts – The Awakening’, a passion project reviving the British philosopher Alan Watts using AI. In an interview with MediaCat’s Content and Social Media Editor, Svilena Keane, Elliott discussed the inspiration behind the project and walked us through its development.
Elliott Starr
Hi, Elliott. Could you please introduce yourself and your work?
Hi, I’m Elliott. I’m a Creative Director working at Impero in London. Prior to this, I worked at a few different agencies. Immediately before Impero, I was at an agency called 20something and I’ve worked at Fallon, Leo Burnett, and Drum as well.
You recently unveiled a passion project using AI to revive the philosopher, Alan Watts. Who is Alan Watts and what was his philosophy?
Alan Watts was a British philosopher who studied philosophy, psychotherapy, psychiatry, Buddhism, Taoism, and Hinduism and had a very good knowledge of history. He lived in California, and he would give these lectures and seminars about questions and challenges that we all face in our lives. I’ve listened to a lot of his work over the years whenever I’ve felt a bit lost in my own life — which I’m sure everyone experiences, certainly in their twenties at least — I’ve just ended up listening to his work with no intention other than just thinking, ‘If I listen to this long enough, I might just find a way through whatever I’m dealing with at the moment’.
He has this amazing ability to elevate you about 10,000 feet above the problem you’re trying to face. I think that’s why his videos on YouTube get so many views; he was an incredibly poetic speaker. He had this ability to just speak off the cuff, unscripted, nonstop. And what you were listening to sounded like it had been written by one of the best screenplay writers in Hollywood.
When thinking about AI, people tend to look to the future rather than the past. How did the idea to revive a philosopher come about and why did you choose Alan Watts?
I think there are many philosophers with lots of published work, but I found Alan Watts’ work really accessible. You could know nothing about philosophy and you could listen to him — I don’t know that that’s true of all philosophers.
The idea to try and bring him back in some form was prompted by two headlines that I read. So one was an article about lonely teens making friends with AI, not getting the social and emotional connection they need in the real world or from real people and seeking that in AI. Then, I’ve been reading other articles around very polarising characters like Andrew Tate; because they’re so polarising and they sit at one end of a spectrum, the binary nature of those people means that they get huge traction on social media since the algorithms effectively decide what’s boosted up. It’s not always about content quality, ethics, morality or basic human decency, it’s just this is being commented on and getting a lot of action. I saw that young people are looking for answers and the danger I see now is that they can just get hoovered up into characters like Andrew Tate, particularly young men.
Young men are so impressionable. They want the world to think they have everything figured out, while in reality, they have very little figured out. The way that they try to pretend and bridge the gap between not knowing but wanting to look like they do is by copying what they see. And now that can happen with someone who lives thousands of miles away that you’ve never met, on social media.
I saw those two things coming together, and I thought there’s so many amazing characters from the past who are just far more positive influences and better role models. I think Alan Watts had a complicated personal life, but through his work, there’s a much better blueprint for a role model than a character like Andrew Tate or any of the clones of Andrew Tate that exist on social media.
Could you walk us through the process of training the AI and the overall development of the project?
I’ve been experimenting with AI probably for about 18 months now and looking at it as a new tool rather than something to be afraid of. I really believed that, with enough time, I could recreate Alan Watts with ChatGPT. That was the first tool that I used.
I worked with a creative at Impero called Beatriz Zambrano, finding essays that Alan Watts had written, sharing them with ChatGPT and then asking it to answer a question and emulate his tone of voice. We’d look at the output and slightly critique it to try and make it sound more like him. I probably spent about two working days just piling thousands and thousands of words into this thread.
There were two things that I did to optimise the thread. I told ChatGPT that I wanted to create a thread where I bring famous British philosopher Alan Watts back to life using AI and I asked it to tell me the best way to word my prompt in order to get the best output. And then as I kept putting more information into the thread, every other prompt, I’d say, ‘Remember, you are Alan Watts. You speak like Alan Watts. You emulate his sentence structure, his word choice, his vocabulary, his mannerisms, terms of phrase, idioms.’
It was a very hard-working ChatGPT thread, but all that gets you is words. So, we took the most viewed YouTube videos of Alan Watts and transcribed those. When we put the transcriptions of the YouTube videos in, they would be entered as: ‘This is best resembling Alan Watts speaking style’. And when we put in any of his essays or his written work, we’d say: ‘This is representative of the way Alan Watts thinks, but not necessarily the way he might speak’.
And then for the output, we thought, well, what questions should we ask? And we’re still asking these, but we just thought, what are the most searched existential questions on Google? We took those as the basis to begin.
We then used ElevenLabs, which is an AI voiceover software. Because this is an art project, it’s not for profit, there’s no client attached and no money being made, that got us away from copyright issues. We provided a sample of Alan Watts speaking and then we got ElevenLabs to generate an AI Alan Watts voice. The first time, it made him sound really American, so we had to tweak the description a bit and say ‘quintessentially British’. In the end, we got a voice that we thought sounded just like him.
So now we had the content, which was the answer to all these questions, and then we had a voice that could answer them. We put those two things together and created these really Lofi videos.
What are your goals for the project and do you have any plans for its future development?
We’ve got no ambition for this project other than to have made it, which we’ve done, and to be useful to people. I think I would have loved something like this when I was coming out of college. And at the moment, it’s a YouTube channel, but if it gets traction, we might release the ChatGPT thread to the public to use. With funding, it could become a totally interactive AI that people could use.
Lastly, what would you say are the key lessons that you’ve learned during the development of this project?
There’s this Japanese practice called Kintsugi — or golden joinery. In the West, people might drop a bowl and it would smash, and they put the pieces in the bin. With Kintsugi, you glue them back together using this mixture of sap and golden powder. So then you have these beautiful golden lines where you’ve glued it back together. The lines highlight the repair. To simplify massively, the idea is the imperfections are what make it beautiful and valuable — it’s stronger for its scars.
You can let the breakage define the bowl or you can let it take on a new life. I teach a fair bit at my old University, and I went up there recently and spoke about Kintsugi and AI. I think, from the projects I’ve done using AI, that’s the attitude we need to take.
Fifteen years ago, people were looking at social media saying similar things, saying things like this is going to break the industry, this is going to ruin the industry and destroy it and all it’s given us is a much bigger canvas creatively to paint on. I look at AI the same way. I just think it’s not going away. It’s here to stay. It’s going to change things dramatically. I don’t agree with just making it your first port of call. If I’m working for a client project to make something beautiful, I want to work with humans to do that. But what AI definitely does is everything that comes before the idea and everything that comes after; you can create so many opportunities with AI. I couldn’t have made this project 10 years ago. I think that would be my general learning from a few projects I’ve done: just embrace it. AI will change things but I think it will just be another part of the creative canvas that we work on.
I think there’s a lot of fear right there around AI, and I don’t think it’s unwarranted. And this is just a personal feeling, but fear is really disempowering. I could sit here and be fearful that AI is going to take my job, but it doesn’t lead to action and because it doesn’t lead to action, it doesn’t lead to output. And creative people just want output; they just want to make stuff. I don’t know if AI will replace me, but I’ll do everything I can to avoid that happening. And in the meantime, I’ll see how much I can increase my output by embracing it.
This interview has been shortened for length and clarity.
Svilena Keane, Content & Social Editor at MediaCat Magazine
Svilena is the Content & Social Media Editor at MediaCat Magazine. She has a joint bachelor’s degree from Royal Holloway University, where she studied Comparative Literature and Art History. During her time at Royal Holloway, she was also the Editor-in-Chief of the student newspaper The Founder. Since then, she has worked at a number of digital and print publications in Bulgaria and the UK, covering a wide range of topics including arts, culture, business and politics. She is also the founder of the online blog Sip of Culture and a self-published poet.
This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.
Strictly Necessary Cookies
Strictly Necessary Cookie should be enabled at all times so that we can save your preferences for cookie settings.
If you disable this cookie, we will not be able to save your preferences. This means that every time you visit this website you will need to enable or disable cookies again.
Performance Cookies
This website uses Google Analytics to collect anonymous information such as the number of visitors to the site, and the most popular pages.
Keeping this cookie enabled helps us to improve our website.
Please enable Strictly Necessary Cookies first so that we can save your preferences!
Targeting Cookies
Cookies that gather information about your browsing habits. They remember that you have visited a website and, in order to provide you with ads that are more relevant, share this information with other organisations, such as advertisers and platforms on which we advertise. Examples: Twitter, LinkedIn and Facebook.
Without these cookies, online advertisements you see will be less relevant to you and your interests.
Please enable Strictly Necessary Cookies first so that we can save your preferences!
On 29 July, Elliott Starr and Beatriz Zambrano announced the launch of ‘Alan Watts – The Awakening’, a passion project reviving the British philosopher Alan Watts using AI. In an interview with MediaCat’s Content and Social Media Editor, Svilena Keane, Elliott discussed the inspiration behind the project and walked us through its development.
Hi, Elliott. Could you please introduce yourself and your work?
Hi, I’m Elliott. I’m a Creative Director working at Impero in London. Prior to this, I worked at a few different agencies. Immediately before Impero, I was at an agency called 20something and I’ve worked at Fallon, Leo Burnett, and Drum as well.
You recently unveiled a passion project using AI to revive the philosopher, Alan Watts. Who is Alan Watts and what was his philosophy?
Alan Watts was a British philosopher who studied philosophy, psychotherapy, psychiatry, Buddhism, Taoism, and Hinduism and had a very good knowledge of history. He lived in California, and he would give these lectures and seminars about questions and challenges that we all face in our lives. I’ve listened to a lot of his work over the years whenever I’ve felt a bit lost in my own life — which I’m sure everyone experiences, certainly in their twenties at least — I’ve just ended up listening to his work with no intention other than just thinking, ‘If I listen to this long enough, I might just find a way through whatever I’m dealing with at the moment’.
He has this amazing ability to elevate you about 10,000 feet above the problem you’re trying to face. I think that’s why his videos on YouTube get so many views; he was an incredibly poetic speaker. He had this ability to just speak off the cuff, unscripted, nonstop. And what you were listening to sounded like it had been written by one of the best screenplay writers in Hollywood.
When thinking about AI, people tend to look to the future rather than the past. How did the idea to revive a philosopher come about and why did you choose Alan Watts?
I think there are many philosophers with lots of published work, but I found Alan Watts’ work really accessible. You could know nothing about philosophy and you could listen to him — I don’t know that that’s true of all philosophers.
The idea to try and bring him back in some form was prompted by two headlines that I read. So one was an article about lonely teens making friends with AI, not getting the social and emotional connection they need in the real world or from real people and seeking that in AI. Then, I’ve been reading other articles around very polarising characters like Andrew Tate; because they’re so polarising and they sit at one end of a spectrum, the binary nature of those people means that they get huge traction on social media since the algorithms effectively decide what’s boosted up. It’s not always about content quality, ethics, morality or basic human decency, it’s just this is being commented on and getting a lot of action. I saw that young people are looking for answers and the danger I see now is that they can just get hoovered up into characters like Andrew Tate, particularly young men.
Young men are so impressionable. They want the world to think they have everything figured out, while in reality, they have very little figured out. The way that they try to pretend and bridge the gap between not knowing but wanting to look like they do is by copying what they see. And now that can happen with someone who lives thousands of miles away that you’ve never met, on social media.
I saw those two things coming together, and I thought there’s so many amazing characters from the past who are just far more positive influences and better role models. I think Alan Watts had a complicated personal life, but through his work, there’s a much better blueprint for a role model than a character like Andrew Tate or any of the clones of Andrew Tate that exist on social media.
Could you walk us through the process of training the AI and the overall development of the project?
I’ve been experimenting with AI probably for about 18 months now and looking at it as a new tool rather than something to be afraid of. I really believed that, with enough time, I could recreate Alan Watts with ChatGPT. That was the first tool that I used.
I worked with a creative at Impero called Beatriz Zambrano, finding essays that Alan Watts had written, sharing them with ChatGPT and then asking it to answer a question and emulate his tone of voice. We’d look at the output and slightly critique it to try and make it sound more like him. I probably spent about two working days just piling thousands and thousands of words into this thread.
There were two things that I did to optimise the thread. I told ChatGPT that I wanted to create a thread where I bring famous British philosopher Alan Watts back to life using AI and I asked it to tell me the best way to word my prompt in order to get the best output. And then as I kept putting more information into the thread, every other prompt, I’d say, ‘Remember, you are Alan Watts. You speak like Alan Watts. You emulate his sentence structure, his word choice, his vocabulary, his mannerisms, terms of phrase, idioms.’
It was a very hard-working ChatGPT thread, but all that gets you is words. So, we took the most viewed YouTube videos of Alan Watts and transcribed those. When we put the transcriptions of the YouTube videos in, they would be entered as: ‘This is best resembling Alan Watts speaking style’. And when we put in any of his essays or his written work, we’d say: ‘This is representative of the way Alan Watts thinks, but not necessarily the way he might speak’.
And then for the output, we thought, well, what questions should we ask? And we’re still asking these, but we just thought, what are the most searched existential questions on Google? We took those as the basis to begin.
We then used ElevenLabs, which is an AI voiceover software. Because this is an art project, it’s not for profit, there’s no client attached and no money being made, that got us away from copyright issues. We provided a sample of Alan Watts speaking and then we got ElevenLabs to generate an AI Alan Watts voice. The first time, it made him sound really American, so we had to tweak the description a bit and say ‘quintessentially British’. In the end, we got a voice that we thought sounded just like him.
So now we had the content, which was the answer to all these questions, and then we had a voice that could answer them. We put those two things together and created these really Lofi videos.
What are your goals for the project and do you have any plans for its future development?
We’ve got no ambition for this project other than to have made it, which we’ve done, and to be useful to people. I think I would have loved something like this when I was coming out of college. And at the moment, it’s a YouTube channel, but if it gets traction, we might release the ChatGPT thread to the public to use. With funding, it could become a totally interactive AI that people could use.
Lastly, what would you say are the key lessons that you’ve learned during the development of this project?
There’s this Japanese practice called Kintsugi — or golden joinery. In the West, people might drop a bowl and it would smash, and they put the pieces in the bin. With Kintsugi, you glue them back together using this mixture of sap and golden powder. So then you have these beautiful golden lines where you’ve glued it back together. The lines highlight the repair. To simplify massively, the idea is the imperfections are what make it beautiful and valuable — it’s stronger for its scars.
You can let the breakage define the bowl or you can let it take on a new life. I teach a fair bit at my old University, and I went up there recently and spoke about Kintsugi and AI. I think, from the projects I’ve done using AI, that’s the attitude we need to take.
Fifteen years ago, people were looking at social media saying similar things, saying things like this is going to break the industry, this is going to ruin the industry and destroy it and all it’s given us is a much bigger canvas creatively to paint on. I look at AI the same way. I just think it’s not going away. It’s here to stay. It’s going to change things dramatically. I don’t agree with just making it your first port of call. If I’m working for a client project to make something beautiful, I want to work with humans to do that. But what AI definitely does is everything that comes before the idea and everything that comes after; you can create so many opportunities with AI. I couldn’t have made this project 10 years ago. I think that would be my general learning from a few projects I’ve done: just embrace it. AI will change things but I think it will just be another part of the creative canvas that we work on.
I think there’s a lot of fear right there around AI, and I don’t think it’s unwarranted. And this is just a personal feeling, but fear is really disempowering. I could sit here and be fearful that AI is going to take my job, but it doesn’t lead to action and because it doesn’t lead to action, it doesn’t lead to output. And creative people just want output; they just want to make stuff. I don’t know if AI will replace me, but I’ll do everything I can to avoid that happening. And in the meantime, I’ll see how much I can increase my output by embracing it.
This interview has been shortened for length and clarity.
Featured image: Alan Watts – The Awakening
Related