It took me 30 years to build a shelf of books and one afternoon to make it redundant

Speaker A: This is the Future of Work. Welcome along to another podcast from Ian and the crew on the Future of Work. Let's join them now for the new episode.

Speaker B: Well, good morning, good afternoon, good evening, Good grief. Welcome to the Future of Work. I, uh, know it says Ian and the crew, but today I'm on my own for an experiment, which is a new form of podcast where it's just me talking. And a happy 4th of July to my American podcast friends. And my actual American friends, too. Normally you'd have Eric and, uh, Maria with me on this, but this week they're off enjoying the holiday weekend. So it's just me, a microphone and an idea I can't stop turning over. I want to talk about AI today, but not the bit everyone's shouting about, because there's a change already happening. Quietly, safely, no drama at all. And it's so quiet that the noise about the robots and the redundancies is drowning it out completely. Here it is. AI has changed what it costs to learn something. And I think that's the revolution that lands first long before any of the ones making the headlines. Now, stick with me, because this one gets quite personal. Here's the shape of the next 15 minutes, two stories, then four moves that you can use tomorrow. And, um, one honest warning. I'm not going to skip first story. The most embarrassing shelf in my house. I used to play competitive bridge. I mean, properly competitive. Before my boys came along, I was ranked in the top 30 for the UK, knocking on the door of the England team. Then the boys arrived and my brain went offline for the best part of a decade. I slid a long way down the rankings. So this year, I decided to climb back. And here's what stopped me dead. Learning bridge now is nothing like learning bridge was 30 years ago. Back then, I bought book after book after book, mining each one for a single golden nugget. That's the shelf I mentioned, by the way. A whole row of them above my desk. Rows of them, actually. Boxes of them. And today they're all an ornament. And I don't know what to do with them, because now I've trained a little AI tutor, uh, on some of the best bridge material. When I hit something I don't understand, I just ask. And I don't ask. What's this bid? I ask the real questions. What is convention actually meant to tell my partner? How often would I, a hand like this even come up? What is the right play here and not the obvious one? The answer comes back in seconds and the Answer that comes back because I asked the right questions. I'll, uh, come to that is the truth rather than just perceived conventional wisdom. And one of the things I've discovered by doing this is that 30 books that all talked about how to do a certain sequence of bids, probably all followed each other and they don't work. So I also trained this AI bridge tutor to do two other things as well. One is I said, make for me a 60 day program to get me to a certain standard. You've got an hour a day of my time. And it came up with a plan of how to do it. Also, I asked it to make for me some distribution patterns. All of that used to take forever to do. The answer comes back in seconds. Compare that with the old way specialist convention, who wrote the best book on it, going to get hold of it. Which of the 300 pages is the answer? On days of hunting for one paragraph. If there's one place I am certain AI changes the world first, it isn't your job. It isn't your week report. It's this. So here's a do this, think of one skill you quietly parked because learning it properly felt too expensive or in time or money. Actually, uh, hold that skill in mind for the rest of this talk. You're going to point today's method straight at it. So before I give you the second story, let me name the insight underneath the first, because it goes far wider than just my card game training. For almost all of history, the bottleneck in learning was never knowledge. The knowledge existed in books, on courses, in people cleverer than us, in professors. The bottleneck was retrieval, the slow, wasteful business of finding the exact thing you need at the moment you needed it in a form you could actually use. You brought the whole book for one chapter. You set the whole course for two ideas. You paid for the entire haystack just to get that one needle. And that's the bit that's gone. I've started calling it a shift. And the shift is from search and hope to ask and receive. And once you've seen it, you cannot unsee it. So that's story one, and it's a nice safe one. Here's story two. This one is actually matters because it happened at work a few months back. I had to get fluent fast in a corner of data governance. I'd spent my whole career politely dodging the kind of topic where everyone in the room has 15 years on you and knows it. Old me would have ordered two books, booked a course for the following month, and turned up at the meeting nodding and hoping. Instead, I did what I'm about to show you. I built a tutor inside our own gated, governed system. Because this is the day job and every skill we use is treated like code. Version controlled, traceable, nothing sensitive, leaking out on an open chatbot. I fed it the article policies, the actual policies, and I spent one evening not memorizing but interrogating it. Why is it done this way? What breaks if we don't? What would a sharp auditor, uh, ask me? The next morning, I walked into that meeting, and I'll be honest, I held my own. Not because I'd become an expert overnight, I hadn't. But because I'd brought myself the right questions and enough of the shape of the thing to know when someone was hand waving. That's the shift in a suit and tie. The cost of getting dangerous enough to be useful in a new domain has collapsed. And for a leader, that's an enormous unlock. So how do you actually put this to work? I have four moves for you today. I'll count you through them one at a time, and each one ends with something you can do today. Move one, narrow it down. A tutor for management is useless, but a tutor for how do I run? Uh, a first one to one with someone who's quietly falling apart is gold. The tighter the question, the sharper the help. Vague in means vague out. For example, with the bridge tutoring, I said I want to be at this competitive level in 60 days. I want to be able to handle the following conventions, and I want to be able to do the following five things as well. Very specific. So move one, narrow it down. Here's your do this today before you next open the tool, or write the tightest version you can. So do this today before you next open the tool, write the tightest version of your question. One situation, one person, one decision. If it barely fits on a sentence, then you're getting close. Right, let's move on. Once you've got your question right, let's move on to move 2. Move 2. Feed it your material. Don't just lean on the model's general knowledge. Train it on the stuff you trust. The book, the notes, the worked examples, the policy. My bridge tutor is only good because I seeded it with real, solid, modern, state of the art bridge material. My governance tutor was only useful because I fed it the real documents and the caution I keep repeating because I live in a regulated world. Only feed it what you're actually allowed to share. Keep anything sensitive inside a properly gated system. Don't hand your company's secrets to an open chatbot to save five minutes because you're saying five minutes and lose your job. Do this this week. Pick one trusted source you already own, a book, a policy, your own notes, and load it in before you ask a single question. Seed first, ask second. So that's moves one and two, a tight question, and your own material behind it. That's the setup. Moves 3 and 4 are where the learning actually happens. Here's move 3. This is where the magic is. It isn't defined this term. It's every question you'd be too embarrassed to ask a human expert to tell you about. For the 12th time, I use phrases like explain it to me, like, I'm smart but new. Why this way? And not the obvious alternative. When does it come up and how often? What would an expert ask me that I haven't thought of? Depth on demand. And nobody's sighing at you across the table. So do this in your next real problem. When you get an answer, don't move on. Ask why this? And, uh, not the obvious alternative again. The second question is probably where the understanding lives. Challenge questions, questions that make you think outside the box, questions that make you struggle. There's a podcast we have on this called Keep being a student who struggles. It's really worth listening to. So that's move three. One more, and then we're done with the method. Here's Move 4. Ask for many, then choose one. When you want examples or ways to remember something, don't ask for one, ask for 25. Keep the three that land. You are the editor of your own education now. That's the job. So do this every time. Stop asking for an example, ask for 25. Bin the duds. Keep what's yours. So there we are, four moves. Narrow it down, feed it in your, uh, own material. Ask the why, ask for many. And then the move that isn't really a move. Go and use it the same day at the table, in the meeting, in the code, because application is what turns an answer into a scar. And the scars are the whole point to get you experienced. Give you one more example of the ask for many. Even this podcast, I fed the transcript to an AI system and said, Give me 20 titles for the podcast and the one it actually has will be possibly one of those 20. Or one of those 20 might make me think of a different angle that I want to work on. And so I will say, I like 16, but do it again with 20 more. Because the ownership of the Output is yours, it's always you. Um, there was a piece of code that was developed with an AI system, uh, in the office that I work. And there was a problem with it. And the person who had commissioned the work from the AI system just sort of said, well, I didn't write it, the AI system did. And we went, no, uh, you supervise someone, you supervise the AI system. You're still responsible for the work. It's your output. It's got to have your name on it. Anyway, four moves now, scale past it yourself. That's the I've given you the personal version. Now let me widen the lens because here's where it stops being a productivity tip and becomes a leadership one. Everything I've just described, you can point at a team. Think about the capability your team keeps renting, the thing that always bring a contractor in for or wait three months to hire for, because nobody internal does that. For most teams there's an obvious one. A slice of data work, a compliance area, a bit of tech stack everyone treats as someone else's job. In the old world, closing that gap meant a training budget, a course calendar, and a quarter of waiting. In this new world though, it's a fortnight. Give four people one adjacent slides, each a gated tutor seeded with the right material, and a fortnight to get dangerous enough to be useful. Then have each one of them teach the others what they found. So this is not just a training program, it's, uh, a learning reflex. And the organization that wins the next decade won't be the ones that automate the most, they'll be the ones that build this learning reflex. So do this in this month. Name the one capability your team keeps outsourcing. Pick one person, give them a gated tutor in two weeks and ask them to come back and teach the rest of you. One capability. One fortnight in house. Now, I promised you one honest warning and here it is. In fact, it's three. Because I'm not here to sell you a miracle. Warning 1. Obviously we know this, it gets things wrong. Of course it does. Think of it as a brilliant, tireless intern with no real world experience, Fast, willing, and every so often confidently wrong. But for learning, that's manageable because learning outcomes come with a built in fact checker reality. Take what is taught to you to the real problem and find out how quickly it works or doesn't work. Warning two from me, the struggle is the point. And it's the bit that you'll be tempted to skip. The reason I can smell hand waving in a governance meeting is 30 years of getting things wrong. If you use this tool to skip the struggle, you'll get the answer, but you'll never get the judgment. Keep being the student who struggles. Let it accelerate the learning. Don't let it replace the learning. Warning 3. And this is the one that keeps me up at night. A young woman, 22 or so once asked me as I was doing a talk at an event. She said this. She said, am I going to get the chance to earn my scars, my experience, before AI uh takes over? It's quite a question to ask, so here's the closest I've got. The tool is a gift. If you use it to ask better questions and go further than you could before, it's a trap. If you use it to stop thinking, the lazy will get lazier, the curious will get faster. Now, that's always true. The leverage is just faster and more enormous. So here's where I land. For 20 years, I taught myself off a wall of books one nugget at a time. That slow, frustrating shelf gave me the single most valuable thing I own. I own and, uh, know how to learn. The quiet revolution is that this skill, once slow, expensive and rare, is now available to anyone with a good question and the discipline to use the answers. So while everyone else argues about which jobs the machines will take, I'd put a smaller, more useful question to you and your team. What could you learn now that you simply couldn't afford to learn before that new starter role? The adjacent discipline, the skill you parted at the start of the episode? Because the organizations that win in the next 10 years won't be the ones that automate it the most. They'll be the ones that learned the fastest. Everyone's now got a tutor. The only question left is whether you'll use yours. So let me leave you now with one move, and it's the four we just walked through, aimed at one target. Take that parked skill point, the four moves at it, narrow seed, ask why. Ask for many and use the answer the same day. One skill, one week. Then, if it works for you, do it for your team. So that's it from me. If you tried and it falls flat on your face, I would genuinely love to hear about it. Next time we head into the Badlands and compare the tools themselves, which model actually to reach for, and, um, all of that. But until then, go and learn something new. For all episodes and the newsletter, come and find me at, uh, futureofwork site or linktree. Ianbanner. Thanks for listening.

Speaker A: That's it for this week. Join us next time for more insight on the future of work. There is also a linktree site which is linktree. Ianbanner if you like this podcast, please like subscribe, follow and tell your friends. Send them a link. It's for free.