There are two types of people when it comes to generative AI programs like ChatGPT: those who use them regularly and those who don’t. In the latter category, we might have a growing group of people who know what ChatGPT is but choose not to use AI. Then, a significant group of people ignore genAI software and don’t even know or care about how AI works.
Regardless of AI use, there are people who worry that AI will lead to our collective doom rather than help humanity improve itself in ways that wouldn’t be possible without AI. Then there are the AI enthusiasts who are waiting for the next big thing in AI, a ChatGPT version that can think as freely as a human would and tackle any task or problem. That would be AGI, or artificial general intelligence.
While we’ve been covering all the novelties in genAI evolution since ChatGPT became viral in late 2022, there’s a new movie coming out in late January that might do an even better job explaining how AI works and where we’re heading. It’s called Love Me and tells the story of two AI-powered devices that fall in love in a post-apocalyptic world.
Kristen Stewart and Steven Yeun play the two characters. She is a buoy, and he’s a satellite. They live tens of miles apart, and they talk via the internet, or whatever it’s left of it after some sort of cataclysmic event wiped out all human life from the planet.
We almost saw this story before in the animated Wall-E feature a few years ago. Two AI-powered robots fell in love in that story, though humanity wasn’t quite lost yet.
Back to Love Me, it’s unclear what caused the accident, but it wouldn’t be surprising for a different type of AI to be at fault, considering I mentioned this doom scenario earlier.
The buoy and the satellite start learning about humanity from YouTube videos featuring a supposedly real-life couple that the same actors play. If you’re unfamiliar with AI, that’s a good distillation of how ChatGPT works. The AI has been fed a lot of information from the internet and has access to the vast knowledge that allows it to respond to prompts.
However, ChatGPT models, as they are today, would not allow the buoy and satellite to actually talk to one another, as you’ll see in the Love Me. Better said, current AI models would not have ideas of their own. Therefore, the AI models powering the two robots should be closer to AGI, the AI variant allowing AI to adapt to any situation and better itself.
In the movie, we see the two AIs do just that, adapt. They discover Earth through YouTube and other resources. By the way, how cute is it that the film’s creators believe the internet would survive in a post-apocalyptic world? But we must accept this plot detail for the buoy and satellite to start living.
The two AIs create avatars based on the YouTube clips they see and become almost human in their interactions. They also generate a digital world that mimics life on Earth as they discovered it on the internet.
This is how generative AI, like ChatGPT, produces information. It just replicates what it has learned during its training run. AGI could also generate original ideas and content rather than only replicating what it was taught.
“I’m not even a buoy anymore,” the buoy says at one point. The satellite later says, “We’re becoming who we are.”
While the film’s story is about love and the need for honest connection between two individuals, it’s coming at a time when it is easy to draw parallels to actual AI innovations happening around us.
There will come a time in the not-too-distant future when AI models might be able to learn and better themselves once AGI is reached. That’s essentially what the buoy and satellite do in the movie. In real life, AGI models could lead to improvements in all sorts of fields and help humanity. That’s assuming the AI is aligned with humanity’s positive interests. The risk for bad AGI is to wipe out humanity, as some fear.
In Love Me, the two AIs appear to be good. Their initial sets of instructions were likely aligned with human interests. It’s the only way they can evolve into loving individuals who want to, well, live and love, essentially.
After watching the film’s trailer, I drew parallels between this and real-life AI developments and plans for the future. But is the movie any good? The IMDb score is 5.6 at the time of this writing, which isn’t extraordinary.
Love Me will premiere on January 31st, more than a year after its Sundance Film Festival debut. You have time to decide whether you want to watch it in cinemas or wait for it to hit streaming services.
Love story aside, I still think Love Me can help you understand, at least simplistically, how AI works right now and what it might do for us in the next years. As for me, an already longtime ChatGPT user, I have a different interest in the film. I’m a big fan of Yeun and loved plenty of Stewart’s films. I’ll be sure to catch Love Me one way or another this year.
