Is This Real or AI?

On this episode I'd like to talk with you about a very small, modern and deeply unsettling moment that more and more of us are having and not quite knowing what to do with. You're looking at a video or a photo or hearing a voice note or a quote attributed to someone. And there's a half second hesitation that wasn't there a few years ago. It's almost like a flicker in your mind of doubt when you asked yourself, is this real?

Did this really happen? Is that actually them? And then you realise that you actually can't tell anymore. And worse, you're not sure that there's any longer a reliable way to find out. I'd like to begin with what that hesitation actually does to a person, because the danger here isn't really any single fake. It's what living in a world of possible fakes does to the inside of your head over time, when you can no longer trust what you see and hear.

Something subtle but enormous starts to erode. And that is your basic unconscious confidence in your own perception. Because for your entire life until the past two or three years, seeing was a kind of bedrock. If you saw it with your own eyes, that settled it. That was the floor you understood. That was the floor you stood your sense of reality on. And now that floor has a very serious crack in it. All the time under everything and you carry this low background uneasiness, a doubt around every image and every claim.

And it's exhausting in a way that's very hard to put our finger on, because it's not the exhaustion of being lied to once. It's the exhaustion of never being able to fully come to rest in anything you encounter. Now the conversation about this, and that's the mainstream one, is almost entirely technical and almost entirely external. And I'd like to talk about it and be fair about it as well, because it's very well meaning.

And it says, get better at detection, learn the tells, look for the weird hands, the strange shadows, the artifacts at the edges, check the source, verify the account, run it through a tool, fact check before you share. And there's some sense in that. But notice the assumption it rests on, because the assumption is already out of date. It assumes the problem is a problem of identifying individual fakes. That if you just got sharper at spotting them, you'd restore your footing.

But the technology is improving faster than our eyes ever will, and the obvious little AI ticks are vanishing, and chasing detection is a race that we are structurally going to lose. And more than that, the detection mindset keeps your attention pointed entirely outward at the screen when the thing that's actually not working that well is something inward. Let me take you to what I think is really happening here, because I think it's deeper and much stranger than deep fakes.

For about two decades, without ever really deciding to, we outsourced our sense of what's real to a screen. We let the feed become our primary organic perception. The main way we found out what was happening in the world, and who said what, and what was true and what mattered. And we took the most fundamental human faculty, there is the faculty of discerning what's real. And we subcontracted that to a glowing rectangle that shows us images, and that worked more or less as long as the images were broadly anchored to something that we can prove happened.

But we built our whole relationship with reality on the reliability of that rectangle, and now that rectangle can manufacture everything seamlessly at scale. So the crisis isn't that the rectangular screen can lie. The crisis is that we handed the screen the job of telling us what was real in the first place. And now the thing we delegated our reality to has become the thing least able to be trusted with it. There's a part of the deep fake panic, almost no one is willing to say out loud.

The deepest danger of AI being able to fake any video or voice isn't that you'll believe something false. It's that you'll stop believing anything at all. Because once you know that the clip of the politician or the photo or the voice note that sounds exactly like your own mother could all be fabricated, the tempting move is to wave the whole world away, to decide nothing can be known. It's all manipulation. Truth is just whatever your side says.

And that shrug is far more useful to a liar than any single fake could ever be. The real prize was never getting you to swallow the lie. It was getting you so exhausted by not knowing that you give up on all knowing at all. Because a person who's decided nothing is real will believe whatever already comforts them. So that's what I want you most to see because it's the one being run on all of us. When everything might be fake, the powerful no longer have to make you believe their version.

They just have to flood the space with enough doubt that you abandon the project of telling truth from falsity. And retreat into believing whatever feels right, whatever your tribe is saying, whatever soothes you. A person drowning in doubt doesn't become a careful, skeptical truth seeker. A person drowning in doubt grabs for the nearest comforting story and holds on for dear life. The collapse of what's real doesn't make people harder to fool.

It just makes them desperate for something to hold. And I'll hold the wrong thing gladly and confidently just to stop the confusion. Now I want to turn this the other way around as well as we always do here. Because there's a second face of the same doubt that no one I don't think is talking about. And it's the one I'd like to sit in the middle of this with you. We've been talking about doubt aimed at the false, the fake video, the invented quote.

But the same suspicion has started landing upon the true. We've reached the point where the moment something looks too good, too coherent, too fast, too much all at once to fit the picture of what a person can do. The first reflex isn't curiosity anymore. It's suspicion. It must be AI. Now, there is a lot of what is called AI slop out there on the internet and on Amazon. And I'm not discounting the fact that that exists.

Of course it does. But one thing I want to talk with you about is that anything that looks like cadence and speed doesn't automatically get reduced to that must be AI. And I want to be honest that this isn't me standing safely outside at looking out at the world and shaking my head at how quickly everyone's become to assume the worst. It happens to me from both directions. It happens to me when people look at my own work and decide that a machine must have made it.

And I'll come to that in a moment. But it also happens in me. I also catch myself doing the very same thing to other people. I'll see something somebody made that's beautiful or prolific. Or maybe it's finished a little too fast and a little too well. And then I feel that ugly little flicker go through me before I can stop it. Did they really do that? Or did a machine do it? So I'm not pointing at the reflex that I'm above.

I'm describing one I'm caught in two. And that's exactly why I think it's worth bringing this into the light. So let me tell you what I've experienced because the honest version is more interesting than the accusation. Over the last year or so I've released under the architect 33 books. And plenty of people took one look and concluded the obvious seeming thing. No one makes 33 books in a year so a machine must have done it.

But here's what those books actually were. 20 years of journals, client notes and clinical notes, my own private journals, workbooks, I'd built over the years for people in programs and courses. Thousands and thousands of pages written one at a time across two decades of doing the work as a therapist and as a man. And when I finally gathered all of it into one place a couple of years ago, I found there weren't a handful of books sitting in there.

There were far more than I'd ever realized. There was a whole coherent body of work from beginning to end. And I spent three months in front of whiteboards in a room mapping out what it actually was. Now I've told the full story of that on a couple of earlier transmissions in the spine so I won't lay it all out again here. And now the part people brace for so I'll just say it. Yes, I used AI but not to write any of the content.

Every word was already mine already lived already on the page. Those books were 20 years of journals, clinical notes and workbooks that I'd already created. One page at a time. What I used the machine for was to compile and synthesize to take that mountain of two decades of material, thousands of pages and help me put each piece where it belonged in the right book in the right order. Now, if I did that manually, that sorting would have taken me by hand the better part of a decade in amongst normal life.

So here's the question I'd hand back to anyone certain that settles it. Why would I spend another 10 years of my one life doing the mechanical work a machine can do in days under my guidance directed by me at every step? When the actual work, the only part that ever mattered, the 20 years of it was already done and written by me. And here's something that unsettles people even more than that so let me push this all the way to the end.

That process didn't just save me a decade of synthesis. It made the books better than I could ever have made them alone. Think about what it takes to hold a single coherent thread across 33 books. One argument, one through line, every idea, every framework consistent across thousands of pages and what has become two million words. No single human mind can hold all of that at once and I certainly couldn't. A man working alone trying to keep one spine straight across 33 volumes, drops threads, contradicts himself, forgets in book 29 what he committed to in book three.

The machine pointed and corrected and directed at every turn by me could hold the whole structure in view in a way that I never could on my own. Now that's not a confession of fraud. It's a glimpse of what the tools actually for. Not to replace the 20 years or to write books for me, but to take the person who lives them and the words he wrote and the frameworks he created and everything that already existed and extend his reach past the limits of one individual's brain.

Which is why is it real or is it AI is already too crude a question for the world we're now in. The life in those books was authored by a person slowly over 20 years. The coherence of them was held by a machine in days under that person's hand, both true at once and the suspicion falls precisely because it wants the answer to be one or the other. And reality has stopped being one or the other. Now this will come for you too in smaller ways.

The work you poured yourself into met with a raised eyebrow, the thing your child made, the thing your friend wrote, the Facebook post, the picture and that little question hanging over all of it. Something devastating happens to a culture when it loses the ability to credit a human being for their own excellence. When it can no longer tell the difference between a person extended by a tool and a person replaced by one.

So discernment, the thing I keep coming back to, was never only a shield to hold up against the current technology that we face today. It's what lets you ask the better question. Is there a life behind this? Instead of the lazy question that everybody's asking now. It's what lets you recognize the real and it's standing in front of you and it's what lets you catch that flicker in yourself the next time you're the one about to wave away something true.

So if the answer to AI fakes was never going to come from the screen and it wasn't because the screen is exactly what broke, then it has to come from the thing we let go. When we started getting our whole sense of reality from a social media feed, an inner compass, not an app that tells you whether the video was real, but a developed human capacity to way, to sit in not knowing without panicking, to feel like a human being.

To feel when a clip is engineered to hit your stomach before your judgment can get a word in. We spent 20 years scrolling letting the phone do that job for us and the muscle wasted away. The work now is to grow it back to become again a person whose sense of what's real doesn't live inside a device that can be made to show you anything at all. And I'd like to be concrete about what that actually looks like because develop discernment can also sound like a slogan.

What it actually is at its core is noticing the feeling something is built to produce in you because the most manipulative content, whether it's fake or real, is almost always engineered to hit you in your unconscious before your judgment can get a word in. The flash of outrage, the surge of fear, the immediate certainty that you already know exactly what this thing means. That speed is the signal. Discernment begins as the simple deliberate act of putting a half second of space between the image and your reaction.

We've talked about the space a lot on this podcast asking what is this trying to make me feel and who benefits if I feel it. It looks like getting comfortable saying I don't know yet and letting that be a stable place to stand rather than emergency to resolve. It looks like learning again to find the real, not only in pixels, but in things that can't be faked. In the people you can actually look in the eye in your own direct experience, in the slow accumulation of trust with a few sources who have earned it over years rather than the fire hose of a feed that shows you everything and lets you verify nothing.

So the most important question in the age of AI isn't how do I tell if this video is real? It's have I built anything inside myself that can sense what's real without the phone's permission? Because the technology will keep getting better at lying, the fake hands and the weird shadows you were taught to look for are already vanishing and your eye will keep losing the race. If your whole sense of reality lives out there in the feed you've already handed it to whoever can generate the most convincing clip.

But if there's a compass inside you, a capacity to wait, to wait, to notice the moment you're being played, to stand steady in the not knowing, then the rising tide of fakes loses most of its power over you. You won't catch everyone, but you'll have stopped needing a screen to tell you what was real. And there's one last layer to this that I'd like to name before we finish because it's the deepest one and it's the one that turns all of this from something happening out there in the culture to something happening in here, in you and in me.

That reflects the must be AI flinch, has slipped below the level we can see. It's now become a function of the unconscious and I've come to think it isn't really about AI at all. It's actually the newest, most efficient form of a very old move, the move we make to discredit something true so we never have to be changed by it. Just think about how that move used to work. When something hit too close, a piece of work, a quote, an idea, a sentence that landed somewhere vulnerable within you.

Somewhere we'd been quietly avoiding. We had ways of making it safe. We'd attacked the person, criticized how they looked, how they talked, the tone they took. We would hunt through the argument until we found one crack, one floor, and use that one crack to throw the whole thing out and it worked. It didn't work by actually answering what was said. It worked because going after the messenger let us stay comfortably blind to the thing inside us that the message had brushed up against.

The thing we always knew somewhere that we should have turned and looked at and never did. Here's what's changed and why it's worse. We don't even have to do that much work anymore. We don't have to read the thing and find the crack to discredit it. We can glance at the cadence, the speed, the articulation of it and say, that's AI. Two words and the whole thing is discounted, unexamined, never even engaged at the surface, let alone, underneath.

The accusation has now become the most sophisticated permission slip we've ever invented for not looking, not at the work and not more to the point at ourselves. We tell ourselves this era has made us sharper, more skeptical, harder to fool. I don't think it has. We've gotten much better at exactly one thing and it isn't telling true from false. It's avoiding ourselves. So that's where I want to leave you because I think it's where the real work actually is, not out there in the fakes but in here.

In the small noticing of the moment you reach for, that's AI or that's not real or look at the state of them. And asking yourself instead whether you're discerning something false or just protecting yourself from something true within you. That pause in the space is the whole game here because the people who come through all of this clear eye won't be the ones with the best fake detector. They'll be the ones who learn to catch themselves in the act of looking away and choose every time to actually look.

If anything in this episode made you want to explore what you just heard, I've made it very easy for you to do so. In the show notes, there's a link to access a book called Before Approaching the Threshold, which is the gateway into this work. Alongside this you'll also receive free 14 day access to the Atlas, an intelligence trained on everything written and recorded. It's there to think alongside you on whatever you're actually sitting with, both are free to start.

And as I said, the link to access them is in the show notes. This was Michael Loria and you're listening to the Architect Speaks.