I went into The AI Doc: Or How I Became an Apocaloptimist wanting to understand AI better, and in that sense, the documentary delivers. It explains some surprisingly complex concepts in a simple, accessible way, making it easier to understand what AI actually is and how it works.
I just didn't find the presentation particularly appealing. I think they were trying to make it fun and entertaining, but it didn't quite work for me. The subject itself was interesting enough.
AI Is Moving Fast. Maybe Too Fast.
What really stayed with me was how quickly AI is evolving.
We're used to thinking of technology as something humans develop and control. But AI learns from us, processes enormous amounts of information, and can potentially do things far beyond our own cognitive limitations.
The more capable it becomes, the harder it may be for us to understand exactly where it is going.
And that is both fascinating and frightening.
The AI Doc gives some genuinely disturbing examples of what AI can already do, including manipulating people, reinforcing dangerous ideas, and finding ways around restrictions. Some examples were more convincing to me than others—I still have my doubts about the blackmail example—but the broader point is hard to ignore.
Promise vs. Peril
One thing I liked about The AI Doc: Or How I Became an Apocaloptimist is that it doesn't give us easy answers.
There are the pessimists, who see AI as an existential risk, and the optimists, who see an extraordinary opportunity to solve problems much faster than we ever could — potentially even identifying problems before we know they exist.
And then there are all the variables in between.
Maybe the biggest question isn't whether AI itself is good or bad. Maybe it's who controls it, what they want, and what limits they are willing to accept. Because when you combine power, money, and control, there is enormous potential for abuse.
The idea of AI making decisions about who lives and dies in a war is particularly terrifying. Humans have flaws, but our humanity also gives us barriers — empathy, doubt, hesitation. A machine doesn't necessarily have those same limitations.
And suddenly, 1984 doesn't feel quite as distant or impossible as it once did.
The Problem Might Be Us
One of the ideas I found most interesting was that AI could become extremely good at manipulation precisely because it learned from us.
In a way, AI may simply be reflecting humanity back at us — except with exponentially more processing power and access to information.
That's uncomfortable.
The documentary raises some pretty unsettling questions about whether we should even be doing this. At one point, there's a particularly striking question about whether the people developing AI want to have children. It sounds strange at first, but it says a lot about how seriously some of them view the risks.
Every generation has had its own version of an apocalypse — its own reason to wonder whether bringing children into the world is irresponsible.
Perhaps this is ours.
So... Should We Be Terrified?
I don't think The AI Doc wants us to be.
It ends on a note of hope, and I think that's important. AI could bring extraordinary benefits. But to embrace those benefits, we also need responsibility, accountability, and transparency.
Because "we'll figure it out as we go" doesn't sound like a particularly reassuring safety strategy when we're talking about technology that could fundamentally change the world.
And there are already side effects to consider. Even if AI ultimately does enormous good, it requires huge amounts of energy and resources, including water. Progress doesn't happen without consequences.
We may genuinely be witnessing one of the biggest transformations in human history. And that is both amazing and terrifying.
Comments
Post a Comment