# GenAI will revive Gatekeeping

On a forum for a monthly contest folks were talking about genAI and someone said "guess what, it's not going away."

(TLDR: The solution is Gatekeeping. Gatekeepers would require an agent or other third-party to sign a notarized document that they've seen/heard earlier versions of the work and can attest under penalty of perjury that it was made by a human. If reader groups require such things then while GenAI fans will be able to publish to Amazon, the readers spending money won't be bothered by the overwhelming mediocrity.)

Guess what? It's easy to make it go away. We don't need the techbros to agree with us. We need the consumers willing to spend money on entertainment to agree with us. That is all. And, honestly, those folks lose with GenAI, too.

It is easy to think the market is artists, authors, and musicians -- but they're really not. The product will never be good enough, (some say it has already peaked), but mostly there aren't enough of them.

It's like Netflix being happy when folks turn on their shows and then only half-watch them. The target market is the completely clueless rando who will pay a monthly fee for a service to get a derivative work where only the AI corporation will profit. They get an endless number of songs, novel sequels, or episodes just like their favorite entertainment from 30 years ago.

Maybe it won't be perfect. Maybe it'll even be glitchy in some places. But it'll be stuff they directly influenced, and that will make it worth the money for them. After all, they aren't artists, writers, or musicians. Even if a team of professionals could always do better, the random person on the street is just a beginner. Using genAI to skip the work -- even if it only gets them 80% there -- certainly feels like a shortcut.

If human musicians, artists, and authors feel like the market is glutted with small independent folks now, it is just beginning because a strong percentage of those folks will see selling this content as a way to get free money.

It's like the gold rush. A few people hit gold and make lots of money. However, the place where the _reliable_ money was found in selling blue jeans and other core supplies to those poor fools hoping to strike it rich. (This is actually where at least one of the long-lasting blue jean companies in the US came from.)

The genAI companies only exist to create profit for genAI companies. They're fine if the artists whose work they're stealing starve because they can't get paying jobs. They're fine if authors' novels are lost in a sea of ChatGPT rivals and new authors who would have had a future instead have none. They're fine if the electricity they need destroys the environment, as long as the toll happens next quarter.

Netflix made enough high-quality shows to sell the idea to people. Now, though, their numbers are better if folks just turn it on and half-ignore it. They want viewers, not awards. It doesn't need to be riveting to be good enough for half-watching.

We're looking at the equivalent to half-watching a Netflix show, but for music and novels. It doesn't need to be perfect. It doesn't even need to be good.

It needs to be good enough to interest the common person. Varied enough, but still in their core strike zone. It just needs to feel worth the money.

I suspect it will result in the end of the era of the lack of gatekeepers. We had a nice little run where anybody could make anything, and their creative endeavors could be seen or heard by others. That's going to be one of the first things genAI ends up killing. Though not, of course, intentionally.

Gatekeeping is the only way to stem the unfathomably large flood of mediocre content that will be coming. Already readers having problems filtering interesting new books they want to read from mediocre genAI stuff that only has a compelling synopsis. (And literary magazines are suffering from the onslaught of genAI short stories.)

Musicians and artists aren't the target consumer for genAI. They're more of a test market. If the process gets good enough for even a fraction of the "real" artists to use it, they can validate the expense and start targeting the real consumer: the average common person who is neither an artist nor a musician but would like to listen to more of a 30-year-old band... but tweaked for their own special interests.

I fully expect future novelists and musicians to need to provide proof that they actually did the work themselves. For a novelist this would be something like a certified document from their agent indicating they had seen earlier versions of the novel and can attest in court that it was made by a human.

It's not that they won't be able to publish it on Amazon without such proof. It's that the gatekeepers protecting the human readers will require it.

I suspect something similar to happen with music as well.

There is some push to make GenAI users be honest about their use of GenAI but... fundamentally GenAI fans are thieves looking to make bank. Telling a little white lie is nothing to someone fine with theft for profit. This is why there are techbros push for technical solutions to "detect" AI writing... but those systems get it wrong regularly, and that's just a cold war as the genAI folks developing systems to pass those checks.

There are solutions that don't require us to trust GenAI fans.

There are solutions that don't require us to wait for techbros and their algorithms.