Get to Know Finn Alexander & The Forever Party's A.I. Generated Music Video

Being wary of the implications of Artificial Intelligence for humanity is nothing new, having been tackled in science fiction countless times ("open the pod bay doors, HAL"), as well as being brought into reality with things like AlphaGo defeating the world’s best Go (sophisticated traditional strategy board game considered to have the most complex move set of any game in existence) or even people’s scepticism towards Alexa and similar software.

Over the last 12 months, the topic of A.I. and its implications for people has never been more real, with the rise to prominence of image/video generators, voice generators and the ever ubiquitous Chat GPT.

Rather than fearing our possibly soon to be A.I. overlords however, many artists are embracing the technology to further their own creativity, rather than merely regurgitate derivative, iterative ideas.

One such musician using these new tools is W.A.’s Finn Alexander & The Forever Party, who have teamed up with artist Kier Spilsbury to craft an A.I. generated music video for their latest single of folk, rock and psychedelic pop sounds, Married to the Machine.

To find out all about the video (premiering on Pilerats today), we picked Finn’s brain - have a watch and listen and get to know below!

Congrats on your new single Married to the Machine! How did this one come together?

Thanks Will! The song started with a handful of lyrics about social media algorithms. I was noting these ‘zoom out’ moments. I’d have set my mind to some task and become suddenly aware that I’d had my head in my phone in this sort of, primitive, societal microcosm for really unhelpful amounts of time. Bingeing on mediocre entertainment, and on peoples online lives. I remember trying to inject a bit of self awareness into the whole thing and realising that for the most part it was sex, and violence, and outrage (and fluffy animals). It isn’t a new thing, media organisations selling us that which targets our primitive minds. But it does feel new and very strange to be a sort of passive receiver, tended to by an algorithm that decides what to feed you for the day.

After I started writing the song, Zuckerberg announced the Metaverse, which will likely distil everything I just mentioned into a sort of ‘liveable’ experience. I think the idea of a world where you can be anonymous and redefined is very attractive. There’s a whole lot of people - maybe most people - that are living terribly unfulfilling lives. I don’t think it’s self-evident that most of us would choose reality over a place where you can perfectly curate your existence. Like all technological advancements, it’s likely that there will be wonderful wins. Information access! Connection! Love! But, those will of course be set against the darker parts of human nature. Things like violence, and pornography, and radicalisation. From that idea the song became a conversation between two friends, both championing different sides of that argument (and both ending up in the same place).

We heard the launch for the single went off - how was it?! How’s the new single feeling live?

It certainly did! The launch was up there with some of the best shows that I’ve ever played, which is always a goal for me. Mojo’s was shoulder to shoulder and the night felt bold and energised and self assured. There was this beautiful circular relationship between us and the crowd and I think the band is finally hitting its stride. It’s taken some experimenting with the setlist to move the crowd in a way that we’re happy with, but we’re getting damn near to what we want. We recently folded in a lead guitarist to add some extra layers and it felt very natural on stage.

In terms of the single, Married to the Machine is a bouncy track and from start to finish one of our more energetic numbers. It’s a lot of fun to play, especially once I get to my ranting towards the end of the song. I’d like to add a synth player for the arpeggiator part at the end though - that part is fucking great. Also, big love to all the wonderful humans that came out, and our star studded supports, Finn Pearson, Claudie Joy and the Joy Boys, Hector Morlet and DJ Bee Rizzi.

The single features a pretty unique music video, being generated by A.I. - how did this idea originate?

Kier Spilsbury, the director (or prompt engineer in the language of AI video generation) is someone who’s work I’ve been following for a while now - she makes these quirky and terrifying horror themed videos using AI diffusion. One of her videos was in the style of early 8-bit video games and so that was what initially peaked my interest. I thought it could work well with the concept of a life lived virtually. The idea sort of gathered mass from there and ended up becoming something completely different, but far more than what I originally imagined it might be. It felt fitting and a little cheeky to write a song critiquing the worship of technological advancement and then have its visual representation almost entirely generated by a computer program.

How did the video come together? Did you find any pros and cons compared to a more “traditional” music video?

It was a lot of trial and error. Kier started by creating what’s called a ‘custom stable diffusion model’ based on images of me. This is pretty much a ‘character’ that can be directed much like an actor would be and inserted into the animations. Kier experimented a lot with different styles and we settled on what is almost a renaissance painting type aesthetic. I then penned a broad narrative arc for the video and sent it through to Kier. After that it was all her. She used a combination of specific prompts ("a young man, person Finn Alexander, is dancing the charleston alone in front of a clean, bright, neon futuristic setting, 60s inspired tokyo future, scifi, soft light, delicate embellishments, painterly, offset printing technique”) and abstract and emotional prompts based on the lyrics ("software's love life, painted on expired film in 2002”). I was pretty much a sounding board from there - “Yes! Phenomenal!” or “Hmmm. Let’s get rid of that.” Then it was very much a matter of piecing together a jigsaw puzzle to best tell the story. The video progresses from the past into the future, so we focused on keeping that consistent throughout. Motion and lighting were also a focus - we wanted the video to feel dynamic!

I think the biggest advantage of making a music video in this way is the room for experimentation that it provides. Going in with curiosity and openness is really helpful - it’s similar to how I like treating songs in a studio setting actually. You let the song dictate what direction you head, rather than the other way around. That’s really difficult with traditional music videos. There’s more constraints when you have to do a physical shoot. I’m not saying AI generation is a preferential method, but it’s a very freeing one. We couldn’t have made a video like this if it required shooting. We just don’t have the resources. A con maybe, of the process, is that it’s hard to get extremely specific. You need to be comfortable with a degree of improvisation as the AI generation process can be pretty unpredictable.

A.I. is obviously a super hot topic across the board right now - what’s your take on A.I. and how it relates to arts/music and artists?

I think perhaps the horse has already bolted when it comes to AI. It’s too useful and too profitable to not to be the ‘apple in the eye’ of market driven tech companies. With that in mind artists are going to have to, albeit begrudgingly, live with it and live with how it changes the creative landscape. I read a book years ago called the Moral Landscape by Sam Harris. He wrote about how AI might come to make art that is as emotionally salient as the art that human beings produce. He talked about how a song, for example, is a series of sounds and words that evoke human emotion. Could one separate a great piece of music out to its components, they might produce a ‘formula’ that could replicate this evoking of emotion.

If that’s what art is then it is feasible that AI could generate art that is indistinguishable from human art. Input ‘love’; input ‘grief’; input ‘the tragedy of Kurt Cobain’. This art would however, be a bastardised version of the human experiment, and ’lesser than’ in the absence of a human creator. Authentic art requires human struggle, and vulnerability, and all the strangeness that accumulates as we walk around being humans. On AI music Nick Cave wrote in the Red Hand Files that ‘algorithms don’t feel. Data doesn’t suffer. ChatGPT has no inner being, it has been nowhere, it has endured nothing, it has not had the audacity to reach beyond its limitations, and hence it doesn’t have the capacity for a shared transcendent experience’. This, I think, should provide a sense of optimism that true musicians will maintain their foothold, and their value. I’m less optimistic when it comes to music that is already excessively formulaic or regurgitated, but perhaps this is not much of a loss. I mean, we mightn’t note the difference.

Look to the writing credits of the ‘top’ commercial pop stars and you’ll notice quickly that many of them serve only as ‘faces’ for the distribution of music that was carefully tailored by ghost writers and marketing teams in the employ of their music labels. They are incredible entertainers; and singers, and dancers, and personalities, but they are not true ‘artists’ (if an artist is to be defined by their ‘making’ of work that communicates their perilous dance with the world). I’m going on here, but there’s a lot to be said. The question of AI art drives us to the core of what art actually is. Art and entertainment are two very different things. Art is brutal and honest and transcendent. While the latter is limbic and sometimes captivating, and without much to say. Art is safe I think, but entertainment should be worried. It may soon find itself in the firing line.

My ranting on music aside, I think that AI generated video should probably be viewed as its own distinctive and compelling medium (perhaps this will happen when AI music evolves as well). The music video that Kier has created is an example of this. While it certainly draws parallels to animation and film making it is distinctively different in its visual aesthetic and the way that it is directed and made. It is not a one-for-one replication of other mediums and I believe it can be an effective tool for artists and for the fast tracked exploration of ideas. I imagine there’s probably vast swathes of revolutionary tools that will come from AI that serve artists to make better art. Optimistically, we should hope that AI furthers our abilities to create, rather than supersede our creating.

What’s next for Finn Alexander and the Forever Party?

It’s pretty busy for us at the moment! We’re heading into RADA studios with Dan Carroll this week to finish off five new songs, we are out in Dwellingup next weekend filming a video for our coming track God is Murdered, and we have a bunch of yet to be announced shows in coming months. The philosophy is to keep making music we love and to keep building on our live shows (I’d love a small string section by the end of the year, but we’ll see how that one goes!) Keep an ear out.

Follow Finn Alexander & The Forever Party: Instagram / Facebook

Follow Kier Spilsbury