When it comes to human civilization, technology is usually the driver, pushing society forward and forward into the future. Fire. The Bronze Age. The printing press. Electricity. The Steam Engine. Radio. Television. Computers. The Internet. All of these had profound impacts on the world we live in. The AI Doc is about the next step in our evolution, artificial intelligence, and what may or may not be awaiting us as this tech we venture into the next horizon of societal intelligence. No T-1000s in sight.
Daniel Roher is a great documentarian, getting an Oscar in 2022 for Navalny. Seeing the bravery of that guy, plus the joy/terror of a new baby on the way, Daniel confronts his biggest anxiety of the moment: what a world of artificial intelligence looks like for humanity when as his baby grows up. Well, when you’re an Oscar Winning documentarian, you can convince CEOs of tech companies like Sam Altman (OpenAI), Demis Hassabis (Google), and Dario Amodei (Anthropic) as well as maybe 50 other people to come try to convince you either way, for just over an hour and a half runtime.
It probably helps that Roher has no technology acumen outside of a camera. The AI Doc is a perfect primer for someone who’s read headlines on the Internet or Social Media bout AI, but has no real understanding of what it is and why people care so much. I’m no expert, but I work enough around it to know the basics, which this doc summarizes quite well. In order to get to his fears, first he has to get us all on the same page with what artificial intelligence is, and how it works, which he succeeds at pretty quickly, maybe 20 minutes in. The concept is really simple: great AI is excellent at pattern recognition, and is fast approaching humanity’s capacity at identifying complex connections, and growing exponentially in intelligence at the same time. He does this using talking heads up and down the field, from the CEO hardcore optimists to the safety officers growing more and more alarmed at the implications of this unchecked intelligence on society.
And then it becomes a debate. Roher starts going back and forth on the key issues of AI: what happens if it replaces most jobs in the world? What happens when we sacrifice checks on AI’s power for the sake of speed? Does the geopolitical landscape make the development risk even greater? Roher thinks the CEO’s will be the big slam dunk of the interview, but anyone who’s worked near the field will know these CEO’s answer the question, well, like, CEOs…non committal, optimistic but cautious, not to implicate themselves. So what Roher shows becomes a personal Rorshach test; if you’re a positive person, you’ll be swayed by the historians and business leaders excited for medical advances to the point of curing most diseases, a real possibility! Or if you’re like me and all the other ex-employees of these places: what’s that gonna matter as history as taught us the poor get screwed everytime with technology changes, and even worse, all those sci-fi dystopias weren’t just fantasy…it could be our future as we put AI into weapons and wars, etc. I don’t share Roher’s apocaloptimist conclusion, because his “what can we do” section relies so much on cooperation of all world leaders I doubt that will happen anytime soon.
But as the documentarian rightly says: what other choice do we have? For the new dad, he would do anything to protect his new baby, and his way to help the world is The AI Doc for starters, so someone else can rise up and feel compelled enough to help make sure this AI transition goes as smoothy for the world as possible. Otherwise, this world’s Skynet is gonna be OpenAI, Deepmind, Grok, or something else.