Hey Weirdos,
He was in the fifth interview. Handshake territory. $475,000 a year. Then he showed the CEO a side project he'd been building on his own time, and the CEO didn't hire him.
He pulled the offer entirely. Not because the work was bad. Because it was too good. Sitting across the table, the CEO realized this guy's technology could erase entire departments of his own company.
That's where episode one starts this week. Episode two starts somewhere stranger: with a man in the UK doing the math on how many hours you've spent, this year alone, essentially having a conversation with a kitchen appliance.
Two guests. Two opposite instincts about the same technology. One says grab it with both hands and use it as a third arm. The other says build a fence around the last patch of ground where you can still prove a human was there.
I think they're both right, and by the end of this email I'll tell you why. Let's get into it.
The $475K Offer He Lost for Being Too Good
At 17, Gabe Campbell was sleeping in a 24-Hour Fitness parking lot. No family. No car. No job. He rode a bicycle to graveyard shifts at a grocery store for $800 a month and taught himself to code out of library books and borrowed laptops. No degree. No bootcamp. No connections.
Fifteen years later he'd built systems for BMW, Google, and Toshiba. He designed the algorithm behind BMW's fast-charge program, the thing that decides when your EV starts charging so the grid doesn't collapse when everyone plugs in at six. He unified design systems at Google for products used by 50 million people a day.
Then came the interview that ended his career as an employee. Gabe walked in and, instead of asking what the boss wanted him to do, he laid out a twenty-year vision, a five-year plan, and a first-year architecture that would have blown up the company's own org chart. As he put it to me, the CEO figured out very quickly that he "was not gonna be a very good employee." Gabe walked away from over half a million dollars in total compensation to bet on himself.
Now he runs AI Genesis with zero employees, using the same agents he sells. His AI replaced 75% of a customer service team at an automotive company and dropped response times from hours to eight seconds.
Here's the part that should make you uncomfortable. I asked him for a truth his industry won't admit, and he said this: it is no longer valuable to move things around on a computer with your mouse. For years, that was a six-figure skill. A great Excel model. A clean slide deck. A beautiful Figma file. "That's gone."
Jobs aren't being eliminated, they're being repriced. Repetitive executors are done. Humans move up into commander roles, and Gabe argues the people who say here's what we should do should now be paid a lot more.
He calls COVID a bait and switch: the pandemic made digital skills mainstream and precious, and six years later AI shrugged and said oopsie, I can do all the computer skills.
The global economy runs at roughly 20% efficiency. In an eight-hour day, most of us are actually productive for one, maybe two hours.
90% of what we've made money doing probably isn't valuable in the next decade.
He's blunt about the risk of universal basic income, too. His fear isn't laziness. It's leverage: after one generation on it, "you now lose the ability to choose, and it becomes tied to compliance."
And yet Gabe is not a doomer. Out of the box, he says, AI is absolutely not your peer. Real peership takes brutal investment: your meeting transcripts, your old emails, your design system, your code library, all of it documented and fed in. He built aiOS to shortcut that, an AI operating system you drive from a text message. His words: "you're inheriting me digitized."
But the line that reframed the whole conversation for me was about what AI actually is. Not a replacement. Not a rival consciousness. A limb.
"I see it as our third arm. I see it as a brain 2.0, a new lobe of the brain. I see it as a suit for consciousness. It's just the new paintbrush."
We went a long way down the rabbit hole after that. Quantum computing. His published Consciousness OS framework on GitHub, ten years in the making, with a twenty-year execution plan and E8 mathematics under the hood. His belief that everything is conscious, that the brain is a receiver and not a generator. And his genuinely chilling aside on quantum breakthroughs: "if I was in the quantum space and I had a real breakthrough, the public probably wouldn't know about me."
He's building all of it in public because he wants to be told he's wrong. "I don't care about being right. I care about being excellent."
Find aiOS and his Consciousness OS framework through his GitHub and the links in the episode description. He reads all his DMs, and he wants receipts.
You've Been Talking to a Toaster for Four Hours a Day
Quick question. When was the last time you were certain the thing you were looking at online was made by a real human being? The photo. The comment under it. The account that liked your post this morning.
71% of images uploaded to social media are AI-generated. Over 50% of long-format text is machine-written. Some platforms are running 45% bots. And according to Scroll Aware, the average 18-year-old alive today will spend 25 years of their life scrolling on a phone.
Ross Thorpe is the founder and CEO of Rooverse, a social platform with one radical rule: zero AI content. Every user is verified as a real human at the door, using the same kind of identity checks your bank runs. Every post is screened by detection software that, yes, ironically uses AI to fight AI. Ross came out of the UK security industry and into higher education, where he watched a generation of students lose their grip on what was real. He's a founding member of an AI trust council.
His origin story is a lockdown story. Young adults in the UK turned to social media as their only form of human interaction, and human interaction is the cornerstone of happiness. But mental health collapsed anyway. Then AI arrived, as he puts it, before Britain had time for a decent cup of tea.
Then he did the arithmetic that stuck with me. People spend four to five hours a day on social media. If you saw someone sitting and talking to their toaster for four to five hours a day, you'd assume they were mad. So what exactly are we doing?
The research he brought is worse than the vibes:
A Cambridge study found that people exposed to a deepfake and then told it was a deepfake had permanently damaged trust in that person or brand. It doesn't snap back to baseline.
MIT and Oxford: just ten minutes of LLM use on a problem-solving task is enough to make you worse at problem-solving, and more likely to give up once it's taken away.
Research out of Japan says LLM use is addictive.
A Nature paper found LLMs can accurately predict the outcomes of social science experiments. Useful for skipping dead ends. Terrifying in the hands of someone asking, "if I made every traffic light red for five extra seconds, would people become more compliant over time?"
And the quiet one: 90% of users post less than once a month. As Ross says, the social has gone out of social media.
So he built a fence. Rooverse has deliberate sign-up friction, metadata and model-level AI detection, and community flagging as the last line of defense, because in security "you have to be right 100% of the time; the attacker only has to be right once." Your feed is just the people you follow, in chronological order. End-to-end encryption stayed, and quantum-safe encryption is going in. And in the most extraordinary insurance policy in social media history, the platform's human-made content is being backed up on the moon, through a partnership with Lonestar Data Holdings, because data centres on Earth get damaged in wars and human-made content is the one thing you cannot regenerate.
He isn't an anti-AI zealot. He wants AI to cure cancer. He just thinks we need somewhere to go.
"People are now living in the AI city, and Rooverse is a nature reserve to go for a walk in the countryside."
Then he handed me the thought experiment I still haven't shaken. You're walking to the meeting that could make you a lot of money, in an expensive new suit. You pass a pond with a drowning child in it. Every single one of you goes in. Ruins the suit, misses the meeting, no hesitation, no calculation. Most days you don't walk past a drowning child. But a child is drowning somewhere. So the question he leaves you with is the one you should ask when you close your eyes tonight: did I make the world a tiny, tiny bit better for being in it today?
Four hours watching the toaster is not the answer. Telling a real artist their real work is good, so they go make more of it, might be. And that's the whole pitch, really. As Ross says, there's nothing stopping people posting pictures of their breakfast on Rooverse. At least you know it's a real person posting a real breakfast.
Find Rooverse at rooverse.app, on the App Store and Google Play. And his own closing advice was the least commercial thing any founder has ever said on this show: if you can choose between talking to a person in real life or going online, pick real life. Every single time. Even over his own app.
The Tool I Actually Use Every Day
Here's something I don't say out loud much: I forget half of what's said in a conversation while I'm still having it. That's why I use Plaud. I backed it back when it was still a Kickstarter, so I've been using this thing since before most people had heard of it.
It records my meetings and calls and turns them into a clean summary with the decisions and action items already pulled out. It's not just the card, either. There's a wristwatch and necklace version I can wear, clip on, or drop on a table, hands-free, so I'm looking at the person instead of my notebook. Best part: I own the notes, I'm present in the room, and the summary is done before I even leave it.
If your brain leads conversations like mine does, grab one here. I've been on this since day one for a reason.
So Which One Is It?
Remember the loop I opened at the top? Two guests, two opposite instincts.
Gabe's answer is to grab the machine and grow a third arm. Ross's answer is to fence off the last acre of verified humanity. It sounds like a contradiction. It isn't. They're both responding to the exact same event: the price of being human is being renegotiated right now, and nobody asked our permission.
If your value came from executing things, Gabe is telling you the market has already moved. If your value came from connecting with people, Ross is telling you the signal is being drowned out by machines wearing human faces. The people who come out of this decade intact are the ones who stop asking what they can produce and start asking what they're actually for — and who then go prove it with results, in public, to real people who can verify it was you.
One builds the arm. One protects the body. You need both.
So tell me: what's the one part of your work you'd hand to a machine tomorrow, and the one part you'd defend to the death? Hit reply and let me know. I read every single one.
Until next week,
Stay weird.
Cody
The Weird Canadian Podcast drops new episodes every week. Subscribe on YouTube so you never miss one.