This video captures a discussion between music pioneer Brian Eno and actor, author, and comedian Stephen Fry discussing their thoughts on artificial intelligence, genital warts and more.
The discussion comes from an industry event focusing on AI-based customer service, which may seem a little off-topic for Synthtopia.
But artificial intelligence is suddenly everywhere and its THE technology of the day. It’s made Nvidia – a company that makes processors used for artificial intelligence applications – the most valuable company in the world. And it’s already become ubiquitous in our everyday lives, for better or worse.
In the world of music, the rise of AI is both fascinating and infuriating.
There are already interesting musical applications of AI, especially where it’s used to empower musicians. But it’s also already clear that tremendous effort is being put into using AI to replace musicians and artists. And it’s already happening.
You can’t do a Google search without getting AI-generated images in your results. And, on services like Spotify, AI-generated music is already showing up in playlists. This is likely to get worse, because services can avoid paying artists royalties by playing more AI-generated songs.
Fry and Eno don’t get too deep into discussing AI-generated music, but the discussion covers a lot of topics relevant to musicians.
One interesting topic that they discuss is how most of the major technology innovations of our lifetimes have had a ‘honeymoon phase’, where the technology is seen as utopian, before we start to understand the true social impact. Fry discusses his initial wide-eyed wonder at the possibilities of Facebook, and how it’s been replaced with a much more skeptical view of the platform and its impact.
It seems that Generative AI art and music may not get much of a honeymoon phase, as it’s already infuriating to many!
Check out the video, and lets us know your thoughts on the sudden rise of AI and it’s impact on the arts.
No one cared about the knitters, coal miners, mechanical draftspeople, lamplighters etc when they were obsolete. Adapt or perish for we still live in a jungle and always will.
Counterpoint: people still listen to music going back centuries.
Too be fair, music hasn’t really moved forward since the early 2000s. AI is only going to add to the mud that really started to pile up after streaming became a thing. Too much output. Too much redundancy that gets in the way of contemporary music what’s actually worth listening to.
Bingo sir. I also noticed the decline starting in the 2000’s right alongside the loudness wars. The other eye opener was when in the last 3 years as I checked out YouTube channels with “musicians” I found out most cannot play and are hobbyists that can noodle around at best. It is a disappointing landscape for sure. It’s no wonder the most popular physical synths every year are little dinky 37 key synths where you cannot seriously compose or play unless it is RAP or EDM.
Nobody’s listening to new music, because we’ve got instant access to all the greatest music ever made.
Loads of people are listening to new music and some of it is really great!
I listen to tons of new music. Much of it I love as much or in some cases more than what I loved in my teens. And 20s. And 30s. And 40s. And 50s. And the music that came before I got here. Doesn’t replace any of it, but just expands it wonderfully.
It does take much more work to find the good stuff now, probably 10x as much as it did in the 80s-00’s, but it’s well worth it to me. I’m definitely not finding it through the idiotic algorithms or Spotify’s “made for you” playlists, which are full of soundalike-but-without-the-depth crap. Independent radio shows, certain YouTube channels, bandcamp, reading blogs, recs from ppl whose opinions I tend to respect. And it’s always fantastic to encounter an inspiring act one has never heard of opening for someone else at a show.
The only downside is feeling out of place as one of the few oldsters at some of my favorite new artists’ shows.
2000s is being generous. I’d say 1990 the future ended as far as music goes.
IDM?
Yeah, that’s arguable. I see IDM as a kind of an experimental offshoot of what was happening in the 80s with techno, which was based on what was happening in the 70s (Kraftwerk, et al).
That’s a broad simplification on my part, I realize, but I see IDM as evolution rather than revolution. Still, some of it was very good, stands up today, and in the early 90s, certianly better than nothing.
To be fair, this sounds like a lovely bit of middle age nostalgia-itis…
Having come of age in the 90’s/00’s myself, I find quite the opposite–music has progressed in fascinating ways, and there are a wealth of popular artists with great chops and remarkable all around musicianship. Much of the best and most boundary pushing music I’ve ever heard has come from 2010 onward from artists who are making a very comfortable living with their music.
I do agree that “too much” has created a mire.
Couldn’t agree more with all of the above. Saying music hasn’t progressed since ’00s is actually saying you’ve fallen out around then and only check what’s on the charts and on top of Youtube searches. Which would give anyone a very dystopian view.
People need not dig deeper than Arca (Mutant), LORN (Drown The Traitor Within), Lucy Plays Wanton Witch, Exploited Body (When My Darkness Came To Light) and numerous others to see that these definitely wouldn’t have been made and have improved on 2000s blog house ffs 😀
There are minimal AI applications for music. It’s nonsense, unless of course you are churning out Sheeran esque shite on an industrial scale (and if so – your career should be eradicated anyhow) there is nothing to fear.
Agree, and synthesizers are just annoying bleepy bloopy machines trying to be real instruments.
People really have to hide that they have used AI to create there music, or when the entire music was created by AI, because when I see either of them I will complete dismiss the music of being of value.
The most valuable and beautiful thing about music is that it was created by a human being.
Making good music is hard and is a skill that needs lots of practice for most people. It is also something that takes a lot of time and effort to make, and is therefor also appreciated by the maker.
A lot of people that lissen to music also make music and a know that these people will feel te same.
Haven’t watched this chat yet, looks like a good way to spend 18 minutes.
At least current-gen AI is limited by human experience & cannot exceed the ability of its designers. That results in the entirely human opportunity to find & break what other humans make.
The future isn’t entirely bleak when people like Benn Jordan are figuring out ways to undermine unethical use of AI:
https://m.youtube.com/watch?v=QVXfcIb3OKo
It’s well known that ai detection is flawed. Reminds me of that new service that protects your art (for money of course)… well it was cracked the next day. Benn Jordan isn’t any different. He’s just riding that youtube controversy-gets-views wave to make a few bucks.
It might help to watch the video I linked. What Jordan was involved with designing was a *_detection_* tool, not a *_protection_* tool.
The program recognises where free AI generated music is being uploaded to services like Spotify. It also accommodates sampling techniques like pitch shifting & layering. My assumption is that the program uses machine learning to achieve wide scale deployment (a nice bit of irony if so).
Of course Youtubers benefit from a certain kinds & degrees of controversy on that platform. That’s how social media functions.
AI generated imagery is now ubiquitous, and it is boring, annoying, and essentially “throw away”. AI generated video has found its way to social media sites like Instagram, and it has become even more tedious and irritating than AI imagery. AI generated music will follow the same path, and to a greater extent than images and video. As mentioned by another reader, it will have no value, and will contribute to the intellectual sewage that defines this strange era in which we are in.
EVERY. SINGLE. GROUP.
You can’t escape people sharing AI-generated crap. “intellectual sewage” is a good way of putting it!
i for one, look forward to our AI music overlords. i am interesterdatgederated in knew keys like G Basil, Z Flatter, and B Musikal.
I think that ai will produce easy music with easy chord progressions and simple harmony and conterpoint , so in any case the music will be mediocre , but the people who will be satisfied with it would anyway compose mediocre music . So the result will be the same . People who compose good music won’t use AI to compose . So no worries anyway .
In generative AI’s most public market face applications, with regard to creative domains, we only get access to a rasterized or baked file.
Prosumer-professional class will have or demand access to layers/stems/metadata for material reworking and modifications of outputs. That will be a distinguishing feature from an Adobe or Steinberg and then an industry feature eventually. Workflow is king and if these slow down workflow they won’t be adopted into industry-grade practices (unless the industry pivots around the technology’s limitations).
Throughout history master painters have painted over the prep work of apprentices and audio engineers have had interns or assistants do their light work, this *could* eventually give such things to everyone once the features are paid for by the investment class for these sorts of technologies.
Great points. The *potential* promise of AI for creatives is offloading grunt work so we can focus our time and energy on higher-level concerns, and your comparison to apprentices doing the prep work is a great way to understand that. One dream is that AI tools can serve for under-resourced artists the role of support team, just as John Williams and Hans Zimmer have an army of grunts to do their orchestrations and other elements and administrative assistants that help them stay focused on the creative work.
But will the best of those tools be accessible to the masses? You seem to question that, and I’m right there with you.
I worry that if AI does fill that dream purpose the side effect will be lost opportunities for apprenticeship and learning under a master, which could lead to a decline of mastery and ability to use the AI tools effectively at all.
If you take pleasure in a piece of art then does it matter how it was created. To suggest a human is somehow the only entity capable of creating true art is a sad thought.
People prefer AI-generated slop to real art, because the average person doesn’t know anything about art:
https://maxread.substack.com/p/people-prefer-ai-art-because-people
This is why musicians are already being replaced. Corporations will make more money, and most people are too stupid to care.
In related news, the world sucks and we are f***ed.
I’d counter with people who like slop don’t care where it comes from and people who appreciate art don’t care where it comes from. Otherwise are we missing the point of art?
“Artificial intelligence” has been evident in the arts for a very long time. It just took the advent of computers to speed it up and give it a simple, public name. I don’t find it very hard to feel the line between genuine art (sincere material with a human touch) and crap designed to sell you widgets. Its sometimes hard to bear, because I truly love my music widgets. Its not their fault that they’re sometimes used to put a poodle jake in a bun and call it a hot dog.