Digital Subversion: How AI Pollutes the Public Well

Simon Carver

[warmly] Welcome to the show. Picture a small town with one well, one water source, and one morning everybody wakes up convinced the water can't be trusted. Not because it dried up, but because somebody polluted it, somebody else lied about who did it, and another person shared the rumor thinking they were helping. That's the world we're in online. Today we're digging into what I’d sum up as the architecture of digital subversion: how synthetic content is turning the public well of trust into toxic sludge. If you like episodes like this, like, share, and subscribe. I’m Simon Carver, here with Lachlan Reed and guest host Jack Burns.

Lachlan Reed

[curious] Yeah, and that well image is bang on. Information used to feel like a shared barbecue in the park. Maybe a bit chaotic, but mostly everyone knew what food was theirs. Now it’s like someone’s swapped the snags for plastic ones and half the crowd is arguing whether fire is even real. Jack Burns, when people say “fake news,” that feels way too floppy for what’s actually going on, hey?

Jack Burns

[calm] It is far too soft a term. “Fake news” suggests a genre problem. This is a systems problem. More precisely, it is a trust problem. And if we want any serious defense, we need distinctions. There are three worth holding onto: misinformation, disinformation, and malinformation. They are not interchangeable. They damage reality in different ways.

Simon Carver

[questioning tone] Okay, give us the clean version. Because I think a lot of listeners hear those three and sort of mentally toss them in one basket.

Jack Burns

Misinformation is false information shared without deliberate intent to harm. Think of a person reposting a bogus claim because they sincerely believe it. It is the accidental litter in the communal well. Disinformation is intentional falsehood, fabricated and distributed to deceive or destabilize. That is not litter. That is poison poured in at night. Malinformation is subtler. It uses real facts, but strips away context so that truth itself becomes a weapon. A leaked message, a real photo, a real event, shown out of sequence or without surrounding facts, can destroy trust just as effectively as an outright lie.

Lachlan Reed

[responds quickly] That “out of sequence” bit is the sneaky one. Because with malinformation, the facts are REAL, but the framing’s cooked. It’s like showing a photo of your mate face-down near the well and saying, “See? He’s drunk again,” when actually he just tripped on a rock carrying a bucket. Same image, totally different meaning. That one sticks with me.

Simon Carver

[reflective] Yeah... and the social damage is almost identical. The town stops trusting the well. People stop trusting each other. And once that starts, the argument isn’t just over one false claim or one clip. The argument becomes whether reality is available at all.

Jack Burns

Exactly. And that is the more serious frame. The objective is often not persuasion in the old sense. It is corrosion. If you can weaken the public’s confidence in shared evidence, institutions, chronology, source integrity—then you do not need everyone to believe one lie. You merely need enough people to doubt everything.

Lachlan Reed

[skeptical] See, this is where I wanna push a bit. Because when people hear this stuff, they go straight to deepfakes. Fake politician videos, cloned voices, the whole circus. But that’s not the whole beast, is it?

Jack Burns

No. That is a common mistake. Deepfakes are only one visible symptom. The broader machinery includes false text, synthetic identities, manipulated context, coordinated amplification, and the deliberate flooding of channels until the human nervous system simply gives up. The goal is cognitive exhaustion.

Simon Carver

[softly] Cognitive exhaustion. That phrase is doing a lot of work. Because it’s not just “Can I spot the fake?” It’s “Can I stay mentally upright while being hit with forty versions of the story at once?”

Lachlan Reed

And mate, forty’s probably generous. [chuckles] Even a kangaroo could trip over this one. You’re not just dodging one dodgy headline anymore. You’re dealing with a whole paddock of them, and some are wrong by accident, some are wrong on purpose, and some are technically true but bent sideways. That’s a rough day out.

Simon Carver

So let’s talk about scale. Because the internet always had rumors, scams, propaganda, hoaxes. What changes when AI walks into the room?

Jack Burns

[matter-of-fact] Scale, speed, and adaptation. In the older model, propaganda required labor. Rooms of people typing, posting, replying, pretending. Now large language models can generate thousands of synthetic personas—distinct names, different writing styles, simulated backstories, even AI-generated portraits that do not trigger a reverse-image search because there is no original image to find. That is an assembly line, not a workshop.

Lachlan Reed

[questioning tone] Thousands of personas. That’s the bit people miss. They imagine one bot account with a weird username and three followers. But you’re saying these things can show up like a whole fake suburb—different accents, different posting habits, different profile pics—all looking normal enough to pass the pub test.

Jack Burns

Yes. And importantly, they can reinforce one another. One account seeds a claim, another “confirms” it, another expresses outrage, another posts a synthetic image, another calls for action. To the casual observer, it resembles consensus. In reality, it may be a coordinated machine performance.

Simon Carver

[curious] A machine performance. That’s good. It reminds me of those improv scenes where three people walk on stage acting like something insane is completely normal, and the audience starts second-guessing itself. Except here the audience is the electorate or the market.

Jack Burns

[calm] Correct. And this is where we should challenge another assumption: the issue is not only “bad actors.” The issue is also the environment that rewards speed over verification, virality over provenance, and emotion over reflection. Bad actors exploit that environment, certainly. But the environment is what gives them leverage.

Lachlan Reed

[interrupts] Provenance. Grab that one. Because I reckon that’s a word people hear and then politely nod like they understood it at school.

Jack Burns

Provenance means the pedigree of the information. Where did this come from? Who created it? What is the chain of custody? Before you engage the content, verify the source. That matters now because the naked eye is no longer a reliable sensor. We have moved beyond the era of obvious AI tells—bad hands, robotic phrasing, the old uncanny valley. Current synthetic media can be good enough that visual intuition is an exploited vulnerability.

Simon Carver

[sharply] “The naked eye is no longer a reliable sensor.” That’s the line I’m taking away. Because for years the public advice was basically, “Just look closely.” But now looking closely is not enough.

Lachlan Reed

Yeah, the old six-fingered-hand trick is dead. Or mostly dead. These days if you’re hunting a fake video, you might need to check metadata, lighting weirdness, tiny language glitches, whether another high-accountability outlet has the same event, same timing, same details. It’s less detective movie, more boring paperwork—which, annoyingly, is often where truth lives.

Jack Burns

[slight pause][skeptical] Boring paperwork is underrated. Triangulation matters. Do multiple independent, credible institutions corroborate this? Does the timing align? Does the source have accountability if wrong? Detection now depends less on “Can I spot the weird pixels?” and more on “Can I verify origin, context, and consistency?”

Simon Carver

And emotionally, that’s hard, because the content designed to fool you usually arrives hot. It shows up already carrying outrage, urgency, humiliation—some big human button.

Jack Burns

That is why emotional decoupling is not self-help language. It is defensive practice. If a post produces immediate fury or panic, pause before sharing. High-impact disinformation is often engineered to bypass deliberate reasoning and trigger a limbic response. In simple terms: if it hits you like a slap, ask who built the hand.

Lachlan Reed

[laughs lightly] If it hits you like a slap, ask who built the hand. That’s grim... but useful. So the practical playbook is pretty plain, even if it’s not sexy. Check the source before the claim. Cool your jets before reacting. Cross-check with more than one trustworthy outlet. And don’t assume your eyeballs are some magic truth machine.

Simon Carver

[warmly] Yeah. And maybe the deeper point from all this is that staying grounded has become an actual skill, not just a personality trait. In an age of synthetic noise, calm is not passivity. It’s discernment.

Jack Burns

[reflective] Precisely. Informational hygiene is now part of civic hygiene. The societies that endure will not be the ones that consume the most information. They will be the ones that preserve the ability to verify, contextualize, and withhold reaction until evidence is sufficient.

Lachlan Reed

[warmly] That’s a solid place to leave it. Protect the well, basically. Don’t gulp from every bucket handed over the fence.

Simon Carver

[warmly] Thanks for joining us, and thanks Jack Burns for helping slow this down in the right way.

Jack Burns

My pleasure.

Simon Carver

If you liked this quick take, like, share, and subscribe. We’ll see you next time on The Human Workforce.