Why are actors afraid of AI? A concrete example

Why are actors afraid of AI?  A concrete example

“The use of artificial intelligence has allowed actors with disabilities to overcome them and be able to acting in films where they could not have actedto work remotely, to rejuvenate and even to get paid forever”: these are the words of Duncan Crabtree-Ireland, executive director of the SAG-Aftra union, which we heard him say in Las Vegas during a meeting precisely on the use of AI in Hollywood.

Six months later, Crabtree-Ireland is the same person and represents the union itself which, precisely because of the risks associated with the use of AI in the world of cinema, thunders against the film majors and proclaims the actors' strike. What happened in these 6 months? Why has Crabtree-Ireland's opinion changed so radically?

The test

From vaccines to war in Ukraine: 10 questions to Bard, Google's generative AI who now speaks Italian

by Emanuele Capone

From Love to Hate: Hollywood and AIs

The most recent thing that happened is that the members of SAG-Aftra were asked by the studios to submit some proposals deemed "inadmissible" (above all, the one on digitization of extras). More importantly, it has happened that in these 6 months the AIs have made enormous strides and are now able to do much more than we could have imagined in January.

Not only aging and deaging (which actors like, as the case of Harrison Ford demonstrates), not only dubbing and lip-sync but much, much more. What is happening is what the writer Nina Schick, who specializes in generative AI, always imagined in Las Vegas, but who at the time passed over in silence: "We will witness a sort of democratization of creativity and entry barriers will be lowered", so that "we will be able to have scripts, stories and special effects at a fraction of the current cost”. And even completely bypassing the people who previously performed these tasks.

That's the biggest fear of the actors and the writers and that's what it would be able to do Showrunner AIdescribed (not by its creators, of course) as “the mother of all Hollywood fears”: unveiled by a company called The Simulation, “it can create episodes of TV series starting from a prompt”, dealing with “screenplay, animation, direction, dubbing and editing”.

The South Park Experiment

Although they are little known in Italy, those of The Simulation are not unprepared: the company was previously called Fable Studio and among its ranks also includes former employees of Oculus and Pixar. Best of all, there's a hands-on demonstration of what Showrunner AI is capable of: coached with hours and hours of watching South Park, the AI ​​managed to create a new 22-minute episode of the hugely popular cartoon on its own. Among other things titled Westland Chronicles and focused precisely on the use of AI in cinema and the writers' strike, in a sort of somewhat curious and quite disturbing loop.

From the company have pointed out several times, too on Twitterto have “no relationship with the producers of South Park” and to have chosen this show for the minimalist style of the animations (currently Showrunner AI is not able to reproduce people in the flesh) and for the enormous amount of material available with which to train the AI.

Another thing immediately made clear is that Showrunner AI has no (at least for the moment) commercial purposes: it was created, as explained in a detailed scientific paperto show "both actors and producers that the threat (of AI, ed) is real" and is deemed so capable that the company does not seem willing to make it public, instead leaving it accessible only to researchers and journalists.

Artificial intelligence

Has ChatGPT gotten dumber? No, but it probably has changed

by Francesco Marino

The ultimate goal: to create conscious AI

It's true: the one made by Showrunner AI with South Parkwhile done really well and definitely believable, it sort of looks like evolution of the so-called fan fiction, however applied to television instead of literature. And yet, according to what has been announced, the potential seems truly remarkable.

For example, what this AI can do is “not only create the episodes of a TV series but also put you inside that TV series”. Put a real person in it, his voice, his somatic features, his way of doing and acting. Which is what Showrunner AI did with Forbes reporter Charlie Finkwho became a lecturer in another (fake) episode of South Park. It's the "15 minutes of fame" of Andy Warhol taken to the nth degree and turned into 6 seasons and 80 episodes of celebrity. Which anyone can be the protagonist of and which anyone could stream on the Internet.

It's not over, why the ambitions of The Simulationaccording to what has been publicly declared, are quite different: not only giving people the possibility of creating their own TV series, possibly being protagonists in them and watching those created by others, but above all setting up a sort of Westworld in the real (albeit virtual) world. Bringing to life a reality show made only by AI, which humans can follow on a daily basis and with which they can also empathize: “Watching Friends we felt like we were really in those friends' lives and we established a connection with them - reads a tweet - Our AI TV will allow see the life of the AI ​​and to establish a connection with them". To do this, from what we understand, other artificial intelligences will be created, similar to Showrunner AI but different: "Our goal is to have AIs that are truly alive, not chatbots that activate only when we talk to them, but AI as people, living their lives true and grow over time”. Theirs the goal is to create an AGI (the acronym stands for Artificial General Intelligence)i.e. an artificial intelligence that is capable of solving general problems by itself and potentially superior to humans.

It is an ambitious goal, it is (was) it same goal as Sam Altman and OpenAI and it is still very early to understand whether it will be achieved or not. What is certain is that the latest case of a show created by an AI and broadcast live online didn't go very well. But that was 5 months ago, and many steps forward have been made since then.


Source link