#25 How to remain irreplaceable in the age of AI

Foreign. Welcome to my podcast, Future of Work, Future Skills and AI. I'm Yasmin Weiss, professor of Artificial Intelligence and the Future of Work. This is no ordinary podcast I created using the AI tool NotebookLM. The characters you hear are two AI voices. However, the content comes from me. Yasmine. Each week you will learn how to work better and smarter with AI. But let's just dive right in. It's what, two in the morning? You're lying in bed, staring at the ceiling, and there's this one question that I think pretty much everyone in the workforce is wrestling with right now. It's February 2026. Hyper automation isn't a sci fi concept anymore. It's just here. And that question that's keeping you up is, in a world where machines can do almost anything, am I obsolete? It's really the defining anxiety of our time, isn't it? I mean, we've gone from the, wow, look at this cool chatbot phase, straight wait, did that software just do my entire week's work in like 4 seconds phase? And that's been. Well, it's been jarring for people totally. And to help us sort of navigate that anxiety and maybe get some sleep, we're doing a deep dive into a brand new piece of work today. It was just published this week, February 3, 2026. It's called Unreplaceable Humanity in the Age of Hyper Automation. And it's by Professor Dr. Yasmin White. And what I really appreciate about this source is that it's not just another think piece, you know, it's not some futurist just guessing what might happen. Professor Weiss is a heavyweight. She really is. She's a professor of AI at work, a leading voice on the future of work. And this is the crucial part. She is in the trenches herself. She's treating her own job as a kind of laboratory. That is the key, isn't it? Yeah. She's not theorizing from some ivory tower. She's actively running experiments on her own workflow to see, you know, where does the human end and the machine begin? So our mission for this deep dive is to get into the mechanics of that. We're going to break down her concept of the substitution matrix, which is a tool you really need to visualize and figure out how to get past the fear of being replaced and into the reality of being augmented. So we're essentially trying to answer, how do you keep your job in 2030 and not just keep it, but maybe even enjoy it? More blunt. I like it. Let's Start with the how. Because before we get to this matrix of doom or hope, I'm not sure which, we need to talk about table tennis. Ping pong. Yes, ping pong. This is the analogy Professor Weiss uses to describe her new work life. She's writing a new book right now, but not like you'd think. Not the lonely author in a cabin. No, no, she's more like a team manager. But her team, well, it's entirely synthetic. She describes the workflow as a game of ping pong. She'll serve the ball, that's a task or a prompt over the net to an AI agent. And the agent hits it back. The agent hits it back. And this is a really important distinction. She's not just, you know, using ChatGPT to write a paragraph. She's treating these things as agentic AI. For anyone listening who's hearing that term, what's the actual difference? It's a huge mindset shift. I mean, a chatbot answers a question, an agent takes on a role, it performs a function. In her experiment, she isn't just prompting, she's delegating. She'll have one agent be the researcher, another one is the critic, another one is the editor. And that's where the ping pong analogy gets really wild sometimes, she says she just steps back and lets the agents play against each other. Exactly. She's orchestrating conflict. She'll have one agent look for data that supports her idea, and then another agent is specifically told to find all the contradictory evidence. She just watches them rally back and forth. And then she decides who won the point, which completely changes her role. She's not the writer anymore, she's the coach. Yeah, the referee. She called it orchestrating, and honestly, it sounds exhausting in a whole new way. She talks about calling timeouts because the agents, they don't get the big picture, right? They don't understand the rules of the game unless you spell them out perfectly. If her prompt is vague, the agent just hits the ball into the net. She has to stop, clarify the rules, explain the context in so much more detail than she would with a human assistant. There's this one part that made me laugh. She said sometimes an agent just plays badly. And she has to, and I'm quoting, send them off the field. It's ruthless. It is. She just deletes it. You're done. Resets it, Programs a new one right there. But there's a really empowering part of this. She makes it clear she's not a coder, she's not a software engineer. Doing this. Yeah, that feels like a big hurdle for people. That feeling of, well, I can't build an AI team. I don't know Python. And she's saying, yes, you can. She uses no code platforms. Microsoft Copilot Suite, OpenAI's GPT Builder, Google Gemini. These are tools where you build an agent using plain English or German. For her, if you can write a sentence, you can build an agent. It's not rocket science, it's just communication. So she builds this incredible machine. Agents are debating, they're researching, they're drafting. And then she hits this psychological wall. The sorcerer's apprentice. Totally. She looks at this thing she's created and asks the big question. By training these agents to be so good at my job, am I making myself obsolete? Why do they need Yasmin Vice anymore? And that's a real fear. If the output is all that matters, then yeah, she's engineered her own redundancy. And that crisis is what led her to develop the substitution matrix. Okay, let's get into the matrix, because I think for everyone listening, this is the part where you need to mentally chart where you are. If you're driving, just try to picture this. It might explain a lot about how you're feeling about your job right now. It's a simple graph, really, but the implications are heavy. So imagine two axes. The vertical one is all about the quality and importance of social interaction. Okay, so how much your job depends on talking to people, Empathy, that kind of thing. Exactly. And the horizontal axis is about the nature of the task. On one end you have optimization and repetition, rules based stuff. On the other end you have intuition and creativity. Social interaction up and down, optimization versus creativity. Left to right. Let's do the quadrants, starting with the scary one. Quadrant one, the danger zone. This is high substitution potential. These are the jobs with low social interaction. And the work itself is very focused on optimization and repetition. Who's on that list? Give us some examples. Professor Weiss lists roles like administrative clerks, quality control, but also, and this one stings for some people, basic software development. Whoa, wait. Coding. We've been told for 20 years that learn to code is the safest career path there is. It was, but think about it. Basic coding is just pure logic. Input A, process B, output C, it's optimization. AI is ridiculously good at that kind of syntax. If your whole job is just turning logic into code without talking to anyone, you're in the red zone. That's sobering. If your job is take data, apply rule, move data, and you do it alone, the Machine can do it faster, cheaper, and it doesn't need a coffee break. It's highly automatable. Okay, let's move to quadrant two. This is where it gets more nuanced. High need for social interaction, but the tasks are still pretty repetitive. She calls this the social anchor. And this is where a ton of white collar professionals are sitting. Think lawyers, Tax consultants, HR business consultants. I feel like lawyers would aggressively disagree that their work is repetitive. Oh, they would. But if you break down the billable hour, a massive chunk is looking up precedents, checking contracts against standard laws. Yeah, that is optimization. It's search. And an AI just destroys a human at that task. So why aren't they in the red zone with the coders, the client, the social part? You don't just pay a lawyer for the document. You pay them to hold your hand, to negotiate, to understand the spirit of the law. That's the social anchor. It holds them in place. So the verdict isn't you're fired. It's more like your job is changing a lot. Exactly. The grunt work, the 80 hours a week of discovery reading for the junior associate, that's gone. But the senior partner, advising the client on strategy, that role stays. It's moderate substitution. Okay, quadrant three. Low social interaction. So you're working alone. But a high need for intuition, creativity. The intuitive anchor. Right. Examples here are criminologists, customs officers, and authors like Professor Weisz herself. Let's stick with the customs officer for a second. Why is that intuitive? Isn't it just checking a passport against a database? Not for the really good ones. I mean, a machine can scan the passport, sure. But a seasoned officer looks at a person and just gets a gut feeling that something's off. They see micro expressions, nervousness, things that aren't in the data. That's human intuition. It's a different kind of pattern recognition. A humankind. An author's fault here, too, which gets back to her whole experiment. The AI can write, but can it create? Can it have that spark of a truly original idea? That's the question. And finally, quadrant four, the safe zone. High social interaction and a high need for intuition and creativity. This is where humanity, for now, really wins. Kindergarten, teachers, elderly care, nurses, doctors in complex situations. It's interesting that these are often the jobs that, you know, we haven't always valued the most financially. It's a total inversion of what we used to think was valuable. I mean, think about a kindergarten teacher. Could a robot teach a toddler the Alphabet? Of course, a tablet can do that now, but do you want a robot teaching your child how to share? Absolutely not. Or how to deal with being sad. In this quadrant, the relationship is the product. We just prefer humans here. You want a human doctor telling you a serious diagnosis. Even if an AI found it, you want that hand on your shoulder. So the big takeaway from the Matrix seems to be that confidence alone is not enough anymore. By 2030, being good with AI will be like knowing how to use Microsoft Word is today. It's just expected. Exactly. If your only value is I process information accurately, you're going to lose. But. And this is the real pivot in her whole argument, don't panic if you just found yourself in that red zone. Yes, this was the part that felt like a lifeline. Yeah. She basically says, you can hack the Matrix, you're not stuck in your quadrant, you can move. And she uses the story of her driver to explain this. It's a perfect example, because on paper, driver is a job that is facing extinction. Right. Autonomous vehicles are here. Getting from A to B is an optimization task. It's deep in the red zone. Technically, yes, a robot can drive a car. But Professor Weiss says she would choose her human driver over a self driving car every single time. And she pays extra for it. The question is why? And it's not because he's a better driver than the computer. It's. It's everywhere. Everything else. He carries her luggage. He asks her how she's feeling because he knows she misses her kids. When she travels, he checks in. He remembers her coffee order. He stalks at Starbucks before he picks her up, so her drink is waiting for her. He makes sure the car has cold water. He's there at the gate. A friendly face, not just a notification on her phone. He took a task driving and he wrapped it in a human experience. He injected humanity into a red zone job. He literally moved himself into the safe zone. Not by driving better, but by being more human. He made the service about the relationship, not just the transportation. And that's the lesson for everyone else. For knowledge workers, it applies directly. Go back to her as an author. Her question was, am I obsolete? And the answer, after thinking about her driver, is no. Because even if the AI writes the draft, she provides what she calls the human varnish. Human varnish. I like that. Sounds expensive. It is. It's the valuable part. It's making sure the book doesn't sound like slick chatbot rhetoric. We've all read that stuff. It's grammatically perfect, but it's soulless. It just glides right off Your brain, it has no texture. Tastes like water. Exactly. She polishes it so it sounds like her. And more than that, people buy the book because they trust her. Yasmin Weiss. They don't want a statistical model's prediction of a sentence. You can't send an AI to a book signing to look a reader in the eye. So she's the soul of the content, the AI is the engine, but she's the one doing the driving, metaphorically, and bringing the coffee. And if we zoom out from that, Weiss has a name for this. She calls it our evolution into Homo augmentus. Homo augmentus, the augmented human. Right. But here's the paradox that she points out, which is so fascinating. All this technology, this hyper automation, it's actually forcing us to become more human in order to survive professionally. That feels completely backward. You'd think technology would make us colder, more robotic, but it's just basic supply and demand. If AI makes raw intelligence and data processing cheap and everywhere, what becomes the scarce, valuable resource? Warmth? Empathy? A real connection. Exactly. The human touch becomes the luxury good. It's the differentiator. She says she's never met a parent who wants a humanoid care robot in their kids daycare. Daycare isn't just about keeping kids alive. It's about teaching them how to be people. She tells that really intense personal story about this on a flight. The lightning strikes. Yeah. Total nightmare fuel. She's on a plane, it gets it by lightning. Huge bang, flash of light, turbulence. Everyone is terrified. And in that moment, the captain comes over the intercom. Now, just imagine if that had been an automated voice. Attention. A lightning strike has been detected. Systems are normal. Oh, that would have been ten times more terrifying. Cold and detached. Instead, it was this calm human voice. The captain explained what happened. He sounded steady. He reassured them. And she said the relief she felt hearing that human empathy, that I'm here with you. I've got. This is something a machine could never, ever replicate. That's the warmth factor. That's the unreplaceable part. And Vice calls this our historic chance. Which is a beautiful way to reframe it. The idea is, if we let automation handle all the boring stuff, the data entry, the scheduling, the first draft, it creates all this free space, and we can fill that space with what actually matters. Human connection. So the future workplace isn't colder. It's cooler technologically, but it's actually warmer because we're spending less time on tasks and more time on each other. A shift from being human doings to. To human beings. That's the hope. So let's bring this all home for everyone listening. We've gone from ping pong with AI agents to this substitution matrix to becoming Homo augmentous. What's the takeaway from Monday morning? I think it's this. Technical competence is necessary, but it's no longer enough. You have to know how to use the tools. You have to be able to play ping pong with the AI. But your job security isn't about being faster than the machine. You will lose that race. It's about being distinct from the machine, doubling down on the things the machine just can't do. Exactly. It's the human varnish, the intuition, the empathy. It's about finding the purely transactional parts of your job, giving those away and using that new free time to be relational. To bring the Starbucks coffee. It really is that simple. Be the driver who cares. Be the captain who speaks with empathy. Be the author with a soul. Which brings us to the end of our deep dive. And based on everything Professor Weiss has laid out, we want to leave you with one final question to think about. It's a kind of self audit. Just look at your to do list for tomorrow. If the doing part of your job, the emails, the reports, the code, if all that was automated away tomorrow morning, how much humanity would actually be left in your role? Are you currently acting like a robot performing a function? Or are you a human who just happens to be using robots? Because that distinction, that's probably what's going to determine if you end up in the red zone or the safe zone. Something to think about. Thanks for joining us for this deep dive. We'll see you next time. If you enjoy the podcast, please subscribe, leave a review and recommend it to your friends, family and colleagues.