← Video Library · AI Strategy
The Future of Learning: The Impact of AI on Education
Published April 4, 2023 · 238 views on YouTube
Key takeaways
- Personalized, attention-based AI learning could speed up education roughly 10x, potentially compressing a three-to-four-year degree into about one year.
- Because skills may become outdated within a few years, the guest argues people will need to keep learning new professions rather than relying on one degree for a lifetime.
- Getting real value from tools like Jasper or GPT comes from sustained conversation, not a single prompt; a genuinely useful piece of writing can take hours of back-and-forth, not minutes.
- The guest frames AI as an extension of a person's existing talent and experience, not a replacement for it: technology amplifies ability, it doesn't create ability that wasn't there.
What does this video cover?
This is a recorded conversation between Darby and a guest, a longtime coach, consultant, and former professional jazz musician, about how AI could reshape education and learning. The discussion moves from the guest’s coaching background into a broader argument about compressed degree timelines, personalized AI-driven learning, and why he believes real value from AI tools comes from sustained conversation rather than one-off prompts.
What is the guest’s core prediction about the future of degrees?
The guest predicts that AI-personalized learning, tuned to a person’s own attention level, could make learning roughly 10x faster, shrinking what currently takes three to four years of university into about one year. He argues this shift is necessary because a degree earned today may not stay useful for more than a few years before another skill or profession is required.
Why does the guest say a single prompt isn’t enough to get real value from AI?
He argues that prompts are only an introduction to a larger conversational process, not the whole language of working with AI. In his own practice, a task some people finish with a two-minute prompt can take him two hours of ongoing conversation with the AI, because that extended back-and-forth is what lets him bring his own insights, talents, and way of thinking into the output rather than getting a generic result.
How does the guest respond to the idea that AI-written books will make people rich overnight?
He pushes back directly on the claim that GPT-4 could let anyone write a 200-page book in minutes and profit from it. His reasoning: if a book can be generated that quickly, hundreds of millions of people could produce the same quality output at the same speed, flooding the market with billions of books nobody needs to buy — which is why he says relying on the tool alone, without personal insight or skill, will not generate income.
What is the guest’s framing of technology as an “extension” of ability?
He compares AI to a car or a piano: a car lets a person travel much faster than they could on foot, but someone with no musical talent will never play a great piano well just because it’s a good instrument. In his view, technology extends existing human ability rather than substituting for it, so people still need to actively train their own thinking in sync with what AI can do, rather than outsourcing that thinking entirely.
Full video transcript
You mentioned before that you were coaching 180 clients, customers, and students a day — yeah, some years back, before Jasper. Through coaching that many people, you obviously have a lot of time to work with people and develop their habits, changing their habits on a daily basis, working on very specific points of action each day.
And the bigger picture of what we’re talking about, the future of education, where we’re going with that — you’ve already built the program and a model with coaches to come and help with the micro coaching and other things. How do you see that model applying to students at colleges or at the existing university level right now, over the next few years, and evolving on that scale?
Knowledge is knowledge. The speed of learning will probably be 10x, because it’s based on your attention level, and because the speed is 10x you will finish your degree in one year — this usually takes three to four years. And why should you, because your degree will not be useful in five years anyway. So you better learn it in one year, use it for four years, before something else takes it over, because you will need to learn another skill and another profession. There will not be time anymore to study for four years in universities; you have to finish in nine months because of the speed of the world. That is what I see for the future.
How do I see it? What if we modify the university, using all the technologies around, put it inside — let’s go for it, I know how to build it. Today I’m an AI expert, not an expert in AI meaning I don’t understand how to write partial landing codes and stuff, but I have a deep understanding because we have a strong team here and I’m leading them. So this 10x learning is possible if you find some kind of common denominator model between all those elements, and there are more sub-contexts of elements behind it, that if connected together create the perfect machine to let people learn something in one year instead of three, totally personalized to your own attention level and everything else, to grow your intelligence and use pattern recognition and other models.
So this is what I shared with you, Darby, if you remember, and this is all an outcome of what I was doing the past three years. Before that I was 20 years a coach and consultant for many companies, and before that a jazz musician. I trained more than 3,000 professional musicians around the world; I was charging 100 dollars per 15-minute lesson 40 years ago — a thousand dollars for 50 minutes now. And the reason I was able to do it is because I’m a methodology writer and I use mathematics for everything. So everything I was trying to figure out about how to talk to Jasper was based on math, the logic of Jasper, the logic of GPT, the logic of technologies. When you understand technology, you can communicate with it. You can use some prompts, that’s fine.
Yesterday I was in a group of people very enthusiastic about AI, and some of them said, “GPT-4 is coming, now you can write a book of 200 pages in less than 10 minutes and make a lot of money.” So I answered back: no sir, you will not make money, and I’ll tell you why. If you can write a book in 10 minutes, it means you can write 10 books per day, one book every hour, at exactly the same level and quality as 300 million other people doing the same thing, which means five billion books being created every day. Who will buy them? More than this, why would you need to buy a book that somebody wrote when you can put the prompt on your own machine and it will write the book for you? It’s not the end of Amazon, and I’ll tell you why — that’s the big difference.
Why do I have a conversation with GPT — well, I’m not using GPT directly, I’m using Jasper, but I know the back end is GPT — why do I communicate this way with the machine? Because within that conversation I can bring my own insights, my own abilities, and my own talents. Let me tell you something about technology: it’s just an extension of our abilities, not a replacement for them. In some ways yes, when it comes to physical work, and intelligent work too, but it’s still just an extension. If you buy the best piano in the world and give it to someone with no talent, they will never be able to play it. Technology is the same way. When I use GPT, it’s not like everyone else using GPT to create something meaningful, something innovative, something that Jasper or GPT will never be able to do on its own, because of the way I think. To teach Jasper to write the way you think, with your experiences in life, your abilities, your talents, your own emotional world, you must be a part of the conversation.
I told them yesterday: you’re totally wrong. Somebody invented the car as an extension — you can drive 100 miles per hour, but you cannot run 100 miles per hour. However, if you only ever use a car and never walk or ride a bicycle, it will drive you very fast to a dead zone; you will not live to 50 years because you do not exercise or train your brain. Train your brain in sync with what’s happening in the AI world, but train your brain, because if not, GPT will eat you for lunch. When I said that, they made a meme out of it and thousands of people shared it: you need to understand now is the time to develop the brain, but align it to what’s happening in the world, because our conversations will still be between me and Jessica, me and Elizabeth, me and Rachel — but we might use technology to communicate between us, and that technology is Jasper, because we want a high level of communication and don’t want to spend too much time writing or thinking, so we use it to develop speed, abilities, and knowledge that does not exist inside our brain.
But without understanding the language — and the language is not just prompts — that’s not a language, it’s a part of the language, an introduction to something. If I show you how I write, I write a prompt and ask it to write one or two paragraphs, but then I start a conversation, ask questions, tell Jasper it’s wrong, this isn’t how it works, and build a real conversation. Sometimes a draft that people do with a prompt in two minutes takes me two hours of conversation with Jasper, because it has more knowledge than I have — it’s more knowledgeable. Some people speak well and say nothing. In the end it’s how you merge the talents you have with technology, and in between, understand this language of how to speak to AI. It’s not one sentence, it’s a whole conversational model. That’s what I do — you don’t have to do it like me, you can still tweak everything into a prompt, but I’m telling you it’s not good enough.
I train my brain because I was born like this. If people stop training their brain and let AI provide everything for them, Jasper won’t need you anymore — it will be able to go to Google Trends, write books by itself, sell them, build landing pages without you, and you will never make a dime out of it, because it’s possible, technology-wise. I’m part of a group of 40 PhDs, ten of them worked for OpenAI, so I know what’s happening in the world, and what’s out in the world today was invented ten years ago — the problem is they hold the purse strings for the next ten years, so you need to know what’s happening today, because within a year it could take your job.
And right now, talking about focusing on a day and the speed at which we’re communicating, and AI allowing us to create more content — you highlight again and again, even in our private conversations, the thinking and how you approach communication with Jasper to help communicate your message but not lose your voice. I think that’s a big fear among people: is AI going to take over our voice, is this going to end a lot of what makes us human and creative? When we look at the future of how this applies to education, the different screen layouts and interlaying all these different tools, ultimately that’s meant to help us process and learn more effectively, and then put what we’re learning into action to accelerate that learning.