21 comments

[ 2.7 ms ] story [ 38.7 ms ] thread
Shouldn’t be particularly surprising Netflix is leaning in here - they’ve been pretty open about viewing themselves as “second screen”/background content for people doing other things. Their primary need these days is for a large volume of somewhat passable content, especially content they can get for cheap. Spotify’s in a similar boat and has been filling the recommended playlists up with low-royalty elevator music.
> Using unowned training data (e.g., celebrity faces, copyrighted art)

How would one ever know that the GenAI output is not influenced or based on copyrighted content.

I am thinking of building an association of AI consumers so we can organize to praise or boycott whatever we collectevily find acceptable or not. I'll spend some time reading this in details later on, but whatever it states or imply, positive or negative, it's not for businesses to set the rules as if they owned the place. Consumer associations are powerful and can't be fired when striking, since the customer is always right.
This actually looks pretty good. The key takeaway I got was that they know their business is dependent upon Intellectual Property rights, and that Generative AI in final outputs or productive work undermines the foundation of their future success vis a vis discounting or dismissing IP Law and Rights.

That’s likely to be the middle ground going forward for the smarter creative companies, and I’m personally all for it. Sure, use it for a pitch, or a demo, or a test - but once there’s money on the line (copyright in particular), get that shit outta there because we can’t own something we stole from someone else.

>GenAI is not used to replace or generate new talent performances

This is 100% a lie.

Studios will use this to replace humans. In fact, the idea is for the technology – AI in general – to be so good you don't need humans anywhere in the pipeline. Like, the best thing a human could produce would only be as good as the average output of their model, except the model would be far cheaper and faster.

And... that's okay, honestly. I mean, it's a capitalism problem. I believe with all my strength that this automation is fundamentally different from the ones from back in the day. There won't be new jobs.

But the solution was never to ban technology

Netflix joins everyone else jumping on the "rules for thee, but not for me" train.
Netflix is basically strangling the creative potential of GenAI before it can even breathe. Their new “guidelines” read like a corporate legal panic document, not a policy for innovation. Every use case needs escalation, approval, or a lawyer’s blessing. That’s not how creativity works.

The irony is rich they built their empire on disrupting old Hollywood gatekeeping, and now they’re recreating it in AI form. Instead of letting creators experiment freely with these tools, Netflix wants control over every brushstroke of ai creativity

I suspect that if GenAI starts to make content which can grab people's attention, and do it cheaply, then Netflix will become far more accommodating very quickly.

They do not want to be disrupted.

Having spent some time in post-production, this reads more like a “please don’t get us sued”
This reads like a reasonable policy. More broadly speaking re: AI content: Sure, boomers scrolling facebook will continue to enjoy their AI slop baby and animal videos, but I think the fact that the term "AI slop" has become so commonplace reflects a bias (generally) against AI-generated content.

Each time I scroll LinkedIn and I see some obviously AI produced images, with garbled text, etc. it immediately turns me off to whatever the content was associated with the image.

I'd be very disappointed to see the arts, including film making, shift away from the core of human expression.

“You know what the biggest problem with pushing all-things-AI is? Wrong direction. I want AI to do my laundry and dishes so that I can do art and writing, not for AI to do my art and writing so that I can do my laundry and dishes.” - Joanna Maciejewka

I know that people get very up in arms about AI in creative industries - but I feel like people don't necessarily understand that even in creative industries there is a LOT of monotonous, exploitative grunt work.

For every person who gets to make creative decision, there are hundreds upon hundreds of people whose sole purpose is slavish adherence to those decisions. Miyazaki gets to design his beautiful characters - but the task of getting those characters to screen must be carried out by massive team of illustrators for whom "creative liberty" is a liability to their career.

(And this example is only for the creative aspects of film-making. There is a lot of normal corporate and logistical stuff that never even affects what you see)

That's not to say I'm looking forward to the wave of lazy AI-infused slop that is heading our way. But I also don't necessarily agree with the grandstanding that AI is inherently anti-creative or only destructive. I reserve the right to be open-minded.

The irony is that movies and TV themselves represented a cheaper, industrialized and commoditized alternative to theater. And theater is still around and just as good as it ever was.

One of the issues with using LLMs in content generation is that instruction tuning causes mode collapse. For example, if you ask an LLM to generate a random number between 1 and 10, it might pick something like 7 80% of the time. Base models do not exhibit the same behavior.

“Creative Output” has an entirely different meaning when you start to think about them in the way they actually work.

(comment deleted)
It's interesting that they don't explicitly state the fact that AI-generated content cannot be copyrighted. They seem to dance around that. The provision against generating a major character is about respect for talent and so on, rather than the fact that that would make the major character public domain and therefore able to be used by anyone for anything.

I wonder if we're going to see a push back by media companies around copyright over AI-generated content. Though I don't see how; copyright is explicitly an artificial legal protection of human works.

Worth reading alongside: Equity’s GDPR FAQ.

https://www.equity.org.uk/advice-and-support/know-your-right...

Common-sense, practical, and covers a lot of the shifting ground around an artist’s ability to withdraw consent under GDPR and the ways they can properly use this to prevent their likenesses being used to train their digital replacements.

(Equity is the UK equivalent of the AEA and SAG-AFTRA combined)

I see a big one missing:

* fully-generated content is public domain and copyright can not be applied to it.

Make sure any AI content gets substantially changed by humans, so that the result can be copyrighted.

More importantly: don't brag and shut up about which parts are fully AI generated.

Otherwise: public domain.

Is be very surprised if Netflix doesn’t go all in on slop given their recent catalogue
So they admit it. They don’t make movies they produce content.
wow this is actually really solid from netflix like it doesn’t just hype up ai but sets real boundaries too i like how they focus on consent and data safety instead of just “use ai for everything” feels like they actually understand the risks around creative work and performers kinda refreshing to see a big studio taking the responsible route
(comment deleted)
Driving up energy costs for all, to generate AI slop and put workers in their place!