Maybe this whole deepfake thing ends with us all going outside
On whether we've ruined something for ourselves that wasn't that great for us anyway
It’s the year 2045. The AI revolution of the ‘20s has settled into a full-blown AI Age, where this technology has insinuated itself into every corner of the digital world. Fortunately, in this future, the worst-case scenarios never happened: there’s no Skynet, we still have a functional economy, the environment hasn’t been irreparably destroyed, etc.
But other nightmares did arrive. AI is in fact able to create music, art, and writing at a quality and scale that humans can’t compete with. Intellectual property is little more than fuel for an unfathomable engine that can replicate anything that exists. Any video you see on the internet could be a deepfake. Anyone you interact with on the internet is probably a bot. No media can be trusted. The well is irreversibly poisoned.
But! We’re in luck. Because our next big invention is, you guessed it, a time machine. And You have been selected to go back in time and warn pre-digital cultures about the dangers of AI.
You step inside the machine, the doors close behind you. After some cool flashing sci-fi lights (which are non-functional but really set the mood), you step out 100 years ago, right during a town hall meeting.
“I come bearing a message from the future!” you say, wowing them with your lights. “You’re not going to be able to trust anything you see on the internet!”
“What’s the internet?” they say.
“Oh, right,” you say. “It’s this thing that lets us all virtually connect with each other. It’s great!”
“That’s amazing!” they say. “I bet it makes you way happier!”
You explain that the general consensus is in fact that it has made everyone really depressed and anxious pretty much since it was invented.
“Oh,” they say.
“Anyway, the point is, the internet is going to be filled with lies and bots and deepfakes so good you can’t distinguish them from reality.”
“Wow,” they say. “That sounds serious. So I won’t even know when I’m on this ‘internet’? It’s like some kind of, to invent a word off the top of my head, ‘matrix’?”
“Well,” you say. “Not exactly. Reality is the same. But digital reality is totally poisoned. Like, you won’t even know if the videos you’re seeing on the internet are real videos or fake videos. For example, any influencer you watch could actually be AI-generated.”
“Oh,” they say.
And then they go back to having meals with their friends and family at tables and only really being aware of local events that actually impact them and working on things that correspond with something real in the world. Almost like they’re living in a post-apocalypse themselves!
Let’s do another little thought experiment. Imagine we’re fish, swimming around in a lake. It’s a nice lake, but after a while, it bores us. So we put our fish noggins together and create a fun little virtual reality to escape to: Lakeworld. You put on a fish-shaped headset and boom, you’re swimming around in a sleek new digital environment.
Lakeworld is fun and convenient. It’s perfect for connecting and communicating. Soon, we’re spending all our free time there. We know it’s not super healthy, but most of us can’t resist.
Lakeworld’s popularity means there’s lots of funding for upgrades. We make it higher resolution and build more fun things to do. And we start designing tools to make our lives there increasingly convenient. One day, we succeed beyond our wildest dreams—we create the ultimate servant. It’s native to the habitat and is perfectly adapted to it. It is a Lakeworld God, omnipotent in the virtual sphere.
Immediately, the Godshark starts eating our little fish avatars and destroying all our Lakeworld infrastructure. And we realize, oops, what we’ve really created is actually the Apex Predator of Lakeworld. Suddenly, nothing we do there matters, and none of us are safe.
Where does this leave us? And might the answer be: right back in the real lake?
If AI rendered a truly Dead Internet, one where distinguishing between fact and fiction, human and machine, was impossible, then what would happen?
You couldn’t get your news online anymore, so it wouldn’t be a 24-hour cycle and probably couldn’t really be on a national or global scale. You couldn’t get social media notifications from friends you haven’t seen in years. You’d have to scroll less because you couldn’t be sure you weren’t just guzzling hallucinations.
Raise your hand if you think that any of these would decrease the quality of your mental health.
Look, I’m obviously being a little flippant about this. The whole deepfake/“synthetic media” thing is unsettling. It’s increasingly impossible to know if what you’re seeing on the internet is real or fake. And we may reach a point where the lines are so blurred that the internet ceases to be of any meaningful use to us.
That would be…uncomfortable. To say the least. But step back a little bit. We’ve shifted paradigms before. And the pre-digital paradigm happens to be the one humans thrived in for millennia. Really, it’s the internet that is the exception, the wild experiment. If it ends with Godsharks, it ends with Godsharks. We had a pretty good run!
The irony of an AI-caused Dead Internet is that all we would have done is render ourselves obsolete in a world that was never ours in the first place. We’re flesh and blood in a world of water and stone. This whole “digital” thing is incredibly useful and very fun, but we were never meant to reside in 1s and 0s, or on screens, or in “the cloud.” Deep down, we know it.
Maybe you think the Lakeworld Godshark is going to find a way to cross over into the real world. I’m deeply skeptical, but that’s a different problem and a question for another day. But assuming that doesn’t happen, then no matter how fake or dangerous the internet becomes, we always have another option in our back pocket: put down the phones and go outside. Who knows, we might like it there.
I know the best way to learn something is to post a wrong answer on the internet, so feel free to educate me. I look forward to being clowned on.
I’ve been saying this for years! Once AI generated content is so widespread as to be undeniable, we’ll have to find institutional arbiters of trustworthiness so that we can reliably verify facts. Probably in a print publication of some kind—let’s say a journal. Perhaps these arbiters would then be called journalists.
I can’t wait for people to return to distrusting what they see on the internet. When I was growing up, I was to never believe what I saw online. And then all the adults forgot!
But I did, Jordan! I went outside for a jog today, because the weather is lovely, and at one point, I nearly bowled over a damn TODDLER because her face was buried in an IPHONE. Her parents didn't even notice, because they were too busy staring down at their phones as well. I thought that touching grass would help me, but it only increased my despair! Who gives an iPhone to a toddler during a family walk on a beautiful day?!
Jokes aside ... you make a great point. Perhaps all this lazy AI-generated content will be a net good after all, if it means we end up spending less time on social media to escape from it. 🙂