I've always hated making album covers. Don't get me wrong - for some musicians, the process of making an album and working with a visual artist is part of the journey - and one that "works" for a lot of people. Just not me. Maybe it's an old-fashioned way of thinking, but I've always believed in the idea of music as its own entity - something you appreciate by listening, rather than by looking.
Although it might sound similar in sentiment, I don't believe in the mantra of "Art for Art Sake", either. (Anything for its own sake eventually destroys itself, always.) But I do believe that the process of listening and looking are two different things that ought to be separated from each other - at least - once in a while.
Even today, my opinion hasn't really changed. "TV/Internet killed the Radio Star" and all that, yes, which I know all too well, having spent most of my adult life in Los Angeles, near Hollywood where all of those sorts of things happen all the time. In recent years, traditional institutions have started relying more and more on image/video content over time, while the internet itself developed a heavy bias towards images/videos while the audio/music layers of it continued to be more neglected over time.
The last time people openly shared music on social media platforms freely was probably MySpace - these days, the internet is largely silent after being litigated anyway on both sides. Influencers and streamers look at music itself with dread, knowing that if they played music they actually liked, they would likely get a copyright strike: de-monetized; if not, outright banned. (Happened to me a few times, too.) People now use music to drown out copyrighted content to try to bypass or confuse the detection algorithms - selected for their disruptiveness, not for quality. Even classical music has been weaponized, used in storefronts to "scare" homeless people away...because classical music is "fancy", and homeless people are not? (This doesn't even work, by the way. Who came up with this dumb idea?)
So it's this sort of insane applications of music that we have normalized in our society today, with seemingly no end in sight. But a lot of it has to do with the fact that we've inserted music into places where they don't belong - and used in ways that are inappropriate and do more harm than good. What we have in most of our apps is a dysphoria engine, designed to make you feel uneasy, paranoid, and anxious about just about everything, including yourself.
"There's gotta be a better way!", you might say. Yes, there is a better way. But some refactoring of how we conceptualize music and art itself is in order - and that will likely require the "separation" between the audible and visual in order to get there.
Content Competes for the Wrong Reasons
Here's where the post starts talking about what's actually in the title: Artificial Intelligence. People who know me probably know my opinion on AI - it's useful for some things, useless for other things - in art, it's a tool that artists can use (or not) but in the end, what separates "good" from "bad" is the talent/effort of the artists themselves.
This is what most "reasonable" people think about AI. In public dialogue, however, there's a polarization of opinions - that all "AI Art is evil/trash" vs "robots will replace all creativity forever" - but the truth is somewhere in between.
I don't really want to get into that middle ground too much because those conversations often turn into subjective discussions about "quality" and "beauty" that have no resolution - instead, we can look at examples where the author clearly doesn't care: people who, like me, loves making music but hates making visuals - especially when they're not necessary.
So yeah, I spent hours and hours learning how to code and making a tool for TEIA where musicians don't even have to think about album covers - just upload the waveform (which is unique and won't trigger copymint checks) and you're done!
While I may be in the minority among musicians, I honestly think that the general public shares my opinion more than people think. Because when you look at all the AI-generated "slop" out there - it's not because the tech itself is bad, but it was clearly a case of where the prompter just didn't care about the outcome itself.
I'm talking about the Tuber who uploaded an AI generated thumbnail that's sort of related to the video, but not really. The well-written article where the cover image is some kitchy thing the prompt came up with that makes the author look shallow, rather than insightful. And of course the numerous "wow, Hollywood is over" posts where the author knows the video is bad but posts it anyway, knowing that it will get engagement from the anger points it generates.
Sometimes less is more - more often than not, I would argue. The reason why "slop" exists isn't because the technology is inherently bad but because the incentive models and data sources it finds itself using has been broken for a while now. Content creators don't have the option to "opt-out" when the algorithms punish you for not "filling in the blank" - so people are pressured into adding "art" into their works where it does more harm than good. The model of "content tuned for the algorithm rather than the quality of content itself" has been accelerated 1000x in the age of AI.
For context, the category of "all recorded music" (including written works) is estimated to be around 15 to 20 million throughout human history. Today, about 1 to 2 million new songs are generated daily, if you include unpublished and amateur works. Album covers are the prime example of this breakdown: since there is no way for humanity to listen to that much music, streaming services are "validating" these songs based on visual content and keywords alone - is anyone even listening to the output? There are people gaming the system out there who probably don't even listen to their own creations - one can assume that much just from the data.
In economics, what happens when a supply curve goes exponential while demand stays the same? The worth of the supply drops to zero, and the whole model collapses in on itself. AI-generated art poses an existential threat to the current model of art production, but it also threatens itself in the long run - because the sheer volume of content risks collapsing under its own weight.
This is not necessarily a bad thing for artists, as long as they keep an eye out for what comes next. If creatives band together and actively pursue alternative models (rather than praying for reform/changes to the status quo which will not arrive), there may be a chance for hitting the reset button and starting anew.
And those changes are happening here, right on TEIA itself, where the dreams of NFTs reforming culture and society at large are still alive and well.
How Artists "Beat" Artificial Intelligence
No matter how you look at it, the business model of the current digital arts landscape is now on the verge of collapse, due to the sheer number of works being produced. In response:
- Some art communities have taken a very strong "anti-AI" stance, hoping that boycotts and legislation will stop the progression towards this collapse. Or going completely "off-grid" and focusing on physical exhibitions and events, and withdrawing from digital outlets as a whole. Inadvertently, however, this becomes a defense of old models (which never treated artists well to begin with) and is a losing battle in the long-run because it cannot be stopped, only slowed.
- Web2 ("Big Data") companies who are reliant on these models have the most to lose, because they are now dealing with a flood of new content being injected into their systems, which is putting strain on their intake methods - already stressed due to the platforms' rise in popularity. Royalty pools were already diluted for the bottom 90% due to overcrowding, and the influx from AI-generated works will likely invalidate the model entirely within a few years. In response, many places have concentrated on "top performers," directing their budgets to a select few - but that comes at a cost: the lack of discoverability and the emergence of new talent, which has led to many of these platforms to become "stale".
- Users on content platforms - social or otherwise - are generally unhappy with the state of things. "Enshittification" is now a mainstream term that people use to describe technology products as a whole, with seemingly no one willing to run counter to those trends.
Web3 (crypto and NFTs) projects leaned hard into the "scarcity" model during their last hype cycle in 2021. There was a zeitgeist-like synergy between the anti-inflationist sentiments of Bitcoin and artists trying to stand out amidst the "noise" of an overabundance of content in the digital world. It was believed that NFTs were not only a source of provenance and protection for creatives, but also a means of quality control - one that would curb algorithmic gaming and empower artists to be part of the conversation.
Then came the memecoins, scams, wallet/DAO hacks, ponzis, the removal of royalties and IP rights, and half-assed implementations of "decentralized" platforms that couldn't even keep the promise of permanent storage - all leading to the rise and fall of FTX, where people's trust in the whole concept disappeared almost overnight. Since then, the markets have never recovered and still struggle to make a basic case for why the average person out there should care.
Then came the next hype cycle - AI - which drained money away from crypto startups and into AI startups, with many of the former pivoting into the latter in desperation. People were tired of the broken promises of decentralization and opted to go in the opposite direction: BigAI, run by huge, centralized data centers operated by big companies we were all familiar with. Sure, they're stealing your data and watching your every move, but at least it was fast and delivered something that felt tangible and new.
At least for a little while. Then the news came out that AI companies were "stealing" creators' data, trying to bypass IP and copyright laws, with media companies suing tech companies in response. The war between the pro-AI and anti-AI camps began to brew - quietly at first, then loudly later. Now everyone finds themselves in the middle of this battle, unsure where all of it is going.
But there is one thing that seemingly everyone overlooked: AI companies typically do not use blockchain data like NFTs in their models because 1) the setup costs for indexing Web3 data require knowledge and expertise that can get expensive, and 2) copyright laws around NFTs lie inside a legal grey area that scares off most big companies before they start.
The reality is that most of the content you see in AI-generated models was "permissioned" by artists who didn't read the fine print, where uploading their work on those companies' platforms automatically grants them display and distribution rights, which has turned into a pipeline for AI models at this point. Even if you find your work "stolen", if you uploaded your work on that social media site that one time, it's very likely there's nothing you can do.
But we're talking about a war over a territory that has been on the decline for years and is literally on a trajectory to implode on itself - is it really worth fighting for? This is one of those cases where the only way to win is not to play.
How to Not Play, How to Win
With both analog and digital art ecosystems struggling to find new directions, future prospects for artists right now may seem very bleak - streaming payments are untransparent and stingy, royalties are not taken seriously and are routinely bypassed, and copyright and IP are not only disrespected but often met with outright hostility. Attribution, too, has become a casualty of the AI era - an artist's style can be ingested, replicated, and redistributed without any credit or compensation, leaving no trace of the original creator behind.
The blockchain solves all of these issues at once - but only if the original visions of provenance, transparency, immutability, and self-sovereignty from the crypto movements are still intact. The watered-down versions of crypto, where people cut corners in exchange for convenience, were never good enough - and the recent shutdowns of the platforms themselves are proof that the markets agree with that as well.
And its within this context where the artists actually have power - the chance to make a difference, for once. If creatives stop feeding into Web2 models and choose to do their work on alternative models like TEIA, even AI companies will be required to negotiate because if they don't have a steady intake of new content being made all the time, they too, will be dead. These companies are too big to be taken head-on - the road to victory is to "starve the beast" by cutting off its food supply. (Art)
"AI Ethics" advocates often argue in favor of building AI models that are more equitable, respect copyright and attribution, and can be held accountable through governance and regulations that ensure their output serves the whole, rather than the few - but how can these ideas be realistically implemented at scale?
Blockchain systems - crypto, governance, ledgers, tokenizations - already have solutions to these problems and can be implemented right away. (Connecting the AI token economy with a crypto token economy is only 1 step away from becoming normalized.) Right now, the reason the two streams don't interact much has less to do with potential and more to do with politics - but at some point, people will realize that the answer was obvious all along: we need to connect AI and crypto systems together to get the positive outcomes people are looking for.
We know that this can already be done. On TEIA, we already have a few things up and running:
- Automated Weekly Reports on TEIA (NoRulesJustFeels): [https://teia.art/objkt/882343]
- AI on NFT Haikus (Joe Simon): [https://8scribo.xyz/chat]
- AI Assisted Art Discovery (FlexasaurusRex): [https://thealcove.art/]
- AI Minting Assistance, Report Generation (Joe Simon): [https://teia-ai-artist.netlify.app/assistant]
Last time I was in Japan, I remember talking about these issues with a few folks there, and the idea of AI systems integrated with crypto was seen as a "no-brainer" - and this was five years ago. While the US struggles to pass basic regulatory laws around Web3 systems, Japan has already gone through the red tape and is paying people directly in stablecoins. With China becoming more involved with open-source projects and pulling way ahead of the West in technological advancement, sometimes I think the race is already lost and we should start preparing to see ourselves as "second place" going forward.
But if there's one thing the West still has an advantage over Asia in, it's that it tends to be better at creating radical new models that defy all predecessors in every way - and this could potentially be one of them, since it's not bound by the stringent rules that often limit approaches in the East.
Whether people come around to seeing it that way is yet to be seen. But to me, the opportunities here are obvious and need to be pursued sooner rather than later. People are dying for situational clarity, and it is needed now more than ever.
Anyway, the whole reason I went down this rabbit hole is still the same as stated in the beginning: I hate making album covers. It's sort of funny how a small preference can define your path so much in the long run, really.