We all know how misleading the media can be. Is this something that we happening with Hollywood concerning the threat of AI?
While independent sources, i.e. non-industry, who look from the outside assert that it is going to send Hollywood into a death roll, the media tends to minimize the impact.
People like Ben Affleck comment on how it isn't going to really affect filmmaking, that movies are protected, and it gets mainstream coverage. On the flipside, when someone like Ashton Kutcher mentions the impact will be worse than most envision, the retorts (which were fierce) are highlighted.
For this reason, perhaps it is best to look outside the epicenter of the industry to see where the threat is. disruption tends to not enter from the incumbents.
Image generated by Ideogram
There is little doubt that every filmmaker is using AI. The only question is to what degree.
India has a large movie industry, one that will likely grow as more tools are available. Considering it is the world's most populous country means that we are dealing with a market where demand is a given.
The International Film Festival of India is steering the conversation in the direction that I wrote about over the last year.
Shekhar Kapur was behind the successful film Mr. India. This is what he had to say about AI.
Kapur said that AI recently helped generate the best plot outline he had received for a potential “Mr. India” sequel. “ChatGPT actually gave me the best plot for ‘Mr. India 2,’ better than most of the big name, well-paid writers,” said Kapur, who directed the 1987 original.
One of the beliefs is that AI will not be able to "create" like humans. This means that scripts have to be written by humans.
Here is evidence, based upon Kapur's statement, that is not the case. ChatGPT produced the best plot for the squeal.
There was a basis for the model to operate from. This is not a new concept since the plot of the original was fed in.
That said, we are seeing the ability to generate something that was previously considered "creativity", thus outside the bounds of AI, is at hand.
My thesis is we are going to see the majority of video content generated by AI within a few years.
Pragya Misra said this about Sora:
Misra demonstrated OpenAI’s text-to-video model Sora, which within seconds can generate minute-long videos from text prompts. While not yet publicly available, the tool is being tested with selected filmmakers. “What you’re going to see today is probably the worst version of the model you’re ever going to see. We’re only going to get better and more intelligent,” Misra said.
We do have to emphasize the fact that Misra works for OpenAi, hence the viewpoint might be biased. The claim made here is accurate though: this is the worst Sora will be.
An event like this is a perfect place to demonstrate the capabilities of Sora. This is the target audience.
Does that mean people will be eliminated completely? Not at all. Instead, it gives someone like Kapur the ability to generate productions without the help of many people.
I stated that future movie stars will be people who do not exist in real life. We are going to see AI hits where the characters become sensations. Due to the ability to generate digital "lives", they will have personas, similar to other actors.
Many feel that AI will not be able to understand (or feel) emotions. I agree with this. It is one of the reasons why I do not fear the Skynet situation. AI will not become human.
So if it doesn't feel, how can AI even produce films?
To me, the answer is simple: it can express it. While it will not know intimately what the feeling it, there is the ability to show the outward expression in film. After all, isn't that what happens?
Take the emotion of fear. An AI does not have a clue what that feels like. However, we can profile what that looks like.
For example, what do people do when they have fear? Depending upon the level, they might shake. Their heartbeat increases. Sometimes they begin to sweat. There is uncertainty. Screaming could take place. Lashing out is common.
The idea here is to picture what someone looks like when fearful. This can be replicated, at some point, by a program such as Sora.
Ultimately, the emotional response is produced within the viewer. This person takes the imagery on screen (along with the audio) and has an emotional response.
In other words, we take the external manifestation of the emotion, as produced by the actor, and respond to that.
AI will be able to do this.
Hollywood is likely aware of this. For the moment, it is mum on what is taking place. Are they blinded to the threat or is it something that is not being talked about?
There is no way to know for sure. What we do know is Indian filmmakers are starting to grasp onto the potential.
We can bet that Hollywood will follow suit.
