I don't know about that, because if I'm interpreting the final clause of the agreement summary, that would give the writer's guild cause for legal action.
That's kind of the problem of trying to figure out what exactly the agreement means from a summary spreadsheet - the final clause smacks of legalese more than the others, so it's hard to know how punitive it is as a normal person.
If it's as punitive as possible, that creates a massive incentive for the companies to keep the writer's guild happy and not set up that parallel production process, which then maintains the guild's power in the next possible strike.
(The question is how long it'll take for that to happen, because LLMs and the processors to run them cheaply and efficiently are going to become more sophisticated and more available as time goes on. So the next strike might not go the guild's way, barring big changes in the legal/regulatory environment, because it'll be faster, cheaper, and easier for the companies to set up an AI based writing pipeline.)
I don't think has ever worked in the history of anything.
Making things harder to acquire/implement through legal and regulatory hurdles? We see that all the time. That's part of how Hollywood has maintained its control of the media (although that control is slipping).
I do think it wise to wait for a more detailed overview, but isofar as what has been released actually summarises the key points, I feel my take on it covers the essentials. Indeed, bans rarely work forever; but they do work for long times, if enough pressure is applied. Usually, this involves the mobilisation of external social drives. (For instance: few now know that the origins of the criminalisation of cannabis stem from the
logging industry, of all things, who saw hemp as a rival source for paper production. The ultimate ban has long out-grown this underlying motivation.)
For now, delaying action is the name of the game. And that works, too. But not for as long, nor as effectively, because better methods tend to have the bottom line to back them up. Case in point: Hollywood may have worked hard to keep control over entertainment media, but as you say: this
is slipping. And that was Hollywood
united, as all involved had a common advantage to defend. On the matter we discuss here, Hollywood is divided, because the interests of the studios are at odds with those of the writers and actors.
As I see it, the writers have lost, because in the long term, this agreement leaves room for the studios to pursue their interests--which they will--to the detriment of the writers. After all, most of these writers offer only mediocre fare at best. It shouldn't be suprising if a well-trained computer can do as well, and do it cheaper.
It wouldn't surprise me, certainly, if a few decades from now most 'mass entertainment' fare is written by machines-- trained by the vast,
vast amounts of source material that are
not owned by the Writers' Guild. The only outcome of this new agreement will be that the writers' guild pushes itself out of the market.
"We refuse to work on these AI productions! We have a RIGHT to refuse! Until you stop using this, NONE of our members will work for you again!"
"That's right. None of you will work for us again. Bye."
Actors may stand in solidarity with their writing "brethren". But by that point, non-actors'-guild-"owned" imprints of faces will be used to train AI, too. And with this, studios will be able to create... not live action, but... life-
like action. Essentially a very refined "animation film" (involving no real actors at all), that
looks perfectly like live-action. It may well be unable to convey profound emotion, but mass entertainment can't do that now, either!
So, in short: I'd find it very plausible, if it came to pass that actual screen-writing and acting
by humans became a form of elite entertainment--like the opera--and considered quite distinct from the AI-created mass entertainment.
deadline.com
