In a post on X (formerly Twitter), Agrim Gupta pointed out that artificial intelligence (AI) systems today are capable of generating 10 times more pixels every two years. Based on his analysis of the rate at which AI-generated video is improving, he concluded that if things continue at this pace, we will have AI-generated TV shows by 2029 and full AI-generated movies by 2031. Services like Sora and Kling have already begun to show us what this might look like.
It seems like it is just a matter of time before we will have no need for human actors. Last year, the screen actors guild in the US went on strike, protesting against the increased use of Generative AI in the film industry. They were concerned that given the rate at which AI is becoming more-and-more deeply integrated with various aspects of filmmaking as a process, it would fundamentally shift the way in which we tell stories through the medium of cinema.
Fears like these are not new. When power looms made textile production faster and cheaper, gangs of out-of-work handloom workers were so incensed that they went about smashing the machines that had displaced them. When banks mechanized the dispensation of cash by introducing automated teller machines, bank workers who were being replaced had similar concerns, though they raised them without resorting to the same level of violence.
Every generation of technology has given rise to similar concerns among the existing workforce—worries that the skills they have amassed over the course of their lives will soon no longer be relevant. And that, as a result, they will become redundant. There seems little doubt that Generative AI will have a similar effect on filmmaking.
Read more on livemint.com