The AI Learning Partner: Socratic Skill Acquisition and Rapid Mastery

The AI Learning Partner: Socratic Skill Acquisition and Rapid Mastery

Raymond Sidney-Smith | 00:00

Welcome back, everybody, to Productivity Cast, the weekly show about all things personal productivity. I’m Ray Sidney-Smith.

Augusto Pinaud | 00:06

And I’m Augusto Pinar.

Francis Wade | 00:07

And I’m Francis.

Raymond | 00:08

Raymond | 00:08

Wade. Welcome, gentlemen, and welcome to our listeners to today’s episode. Today, we’re going to be continuing our AI-powered professional series, And in this particular episode, we’re going to focus in on how AI is… Helping people become better learners. Basically, how these tools are transforming personal and professional development. And so we’re going to be talking about passive consumption, how we use maybe some Socratic method, and how we can do all of this through the creation of AI acting really as a personalized mentor tool. To basically challenge our understanding, accelerate our learning curve, and help us achieve proficiency, hopefully faster than before. And so let’s talk about what traditional learning really looks like, and then we’ll Talk about what the… Problems are in that space that we’re trying to solve for when we’re utilizing AI.

Francis | 01:09

So the traditional model is that you sit down and receive knowledge from someone else. That there’s an expert. Behind the scenes. But what’s happening now is way more fluid than that. It’s that. I’m somehow reaching out to an entity and it’s reaching out back to me in real time. And it’s an AI entity. 

So no experience is give and take between you Something that’s really smart. And where I am. And this give and take is not Obviously, it’s not one directional, but It changes over time because I’ve noticed that cloud for example, is becoming better at finding stuff for me or teaching me stuff. 

So it’s like having this flexible space in which I’m being exposed to more things, I’m learning more things, I’m remembering more things. Because it’s reminding me of more things. I’m accessing more options. And in so doing, I’m becoming a better professional. In my case, I’m thinking in particular of writing. I’m becoming a better writer. It couldn’t have happened with a human being. The way it’s happening with AI. And for the last year and a half, I’m really grateful for it because this kind of flexible space is Absolutely making me a better professional. 

So it’s again, It doesn’t compare to anything that I’ve experienced before. So it’s something that’s brand new and it’s evolving. I.

Raymond | 02:38

Do, too.

Augusto | 02:39

Hello. It is interesting. I’m going to talk about two cases that I recommend. For social media for artificial intelligence. One is I sometimes work with clients who have beaten up. Okay, beaten up as a child, beaten up as an early adult, beaten up. And one of the things that AI produce a different experience. Delusional, but a different experience is that AI will try to make you feel good. And I have recommended to a couple of my clients, go and have a glass of wine and a chat with your chat GPT. Okay, feel good. Because Feeling good, it’s very important for productivity. It’s very important for learning. It’s very important. If you go, okay, I’m going to study this book. And you get into that, I’m going to study the book. But then you go like, okay, well… I don’t know how to learn or I can’t get into this material. The learning, the growing is very hard. Feel nice. 

So that’s one thing. But the second is how much Really? You can adventure. Into new topics, even if you don’t want to master them. Okay. I remember Good day. Before Google, yes, because I’m that old. Okay. Where if you wanted to repair your car, okay, your best hope was to go to AutoZone or the equivalent and find a Hanes manual. Okay. And that thing was like the golden source. You could have every screw of that car had was on that manual. It was incredible. Then came YouTube and we asked how we did it before YouTube where you can see the video. Now it’s even better. I have a chat for every vehicle we have in our house. Okay? And I can go and ask, not for, okay, how do you fix, no, for this specific model of this specific year of this. And it saves even more time. I enjoy a glass of wine once in a while. I’m very bad at it. Okay? According to a friend of mine, the reason is I don’t drink enough wine. Okay? And if you get very good at it, you need to drink enough wine. I fail. Okay. But this is what is interesting. I’ve been putting into ChatGPT, okay, the wines that I enjoy. Okay. When I, and, What happened the other night, I was having dinner with some friends. We were in their house, and a friend of mine pulled three bottles of wine. I said, pick one. In the past, I will have picked based on the label. Guilty of charge. I know nothing about it. This time, I pull my phone, take a screenshot or a picture of the three labels and say, what matched my profile? And it picked something completely different to what I was going to pick. But It was actually fitting to the profile. It was a glass of wine that I enjoy because in the past what happened is I get the glass of wine and No, this one was. You can go from the mundane And you can really go with this. To the expert level. I was having a work issue with a VDI. For the people who don’t know what a VDI is, it’s a virtual desktop interface. And I basically have a client who gives me this VDI to do work with them. Fantastic. Except that I could not make it work for what I wanted to make it work. Chat GPT to the rescue? No. Okay, I went put all the details blah and it says the problem is you are trying it was an audio issue you are trying to make This on your iPad. The solution is try to remove I remove one of the layers it worked better and then I went and bought a And they’re flying. Problem solved. The problem is the machine and the VDI cannot go and convert the audio to the VDI audio, to the iPad audio, to Bluetooth. Great. And remove two of those layers and problem solved. But these were things that in order to get to those conclusions, the number of trial and errors And the time that you need was… Among us. And now you can get those answers with two or three good prompts very quick. That bring me to the last point I’m going to make. I remember the first time I read Alan Toffer sentence saying, The analfabets of the future will be those who are unable to learn and unlearn. If he will be alive today, he will say that alphabets of the future will be those who can make prompts. And this is something when people say, then AI is going to take our work and the world and all this. Yes, it’s going to change it. No question about it. And Terminator will be in the right of the corner. That said. The right prompt will really make this tool A tool? And if you don’t learn how to do prompt and just think you can do what you did with Google, You are. Much likely dead in the water.

Raymond | 07:48

There are a couple of things that I think AI can help with here in terms of learning. When you have full agency over your life, the way in which you learn is going to be different. When you have an employer who may be asking you to learn as you onboard into a new company, you may have to watch a whole bunch of learning videos to onboard into the company. That is a fundamentally different kind of learning than a teacher teaching you, a professor teaching you, a topic so that you can get a degree and move on in the education system. 

So we need to think about the way in which we are thinking about learning as adults and that I think that perspective alone changes when you start approaching AI as what I talk about as basically being a learning partner. Because this becomes not just, I mean, there’s so many different categories. I can go in a lot of different directions, but I will, I’ll say that one, just learning on the fly, it becomes this wonderful learning experience. Sidekick, a learning partner, but it helps you identify things where you might need certifications or other kinds of ongoing continuing education in your career. And it can help you in those regards. And it can help you with lifelong learning, just having that joie de vivre, wanting to engage in life in a way that it’s meaningful. And I think that there are a couple of major problems in learning that we want to challenge. And I think that’s the foundation of some of this learning piece. And again, I think there’s some missing pieces. Approaches here, but I think generally we know that we learn best from when we are quickly assessed on the topic that we’re trying to learn. 

So testing or self-assessment really helps the brain quickly identify where it has gaps. And so you need to have that kind of self-assessment, that kind of personalized feedback that AI can provide over and over again. And it doesn’t get tired. Different than if you had, say, a coach or a professor or a trainer sitting in front of you for hours on end, they would get tired. Mentally fatigued, they can have a bad day. Any number of things. The AI doesn’t have that. You could curse at the AI and it’ll just like shrug it off and just continue being nice to you, which happens when you’re learning as an adult, right? You can get frustrated and different than a child. You probably don’t have agency to just get up and walk out of the room as a child from a classroom. As an adult, you can just get up and walk out. You’re a grown person. You can leave. And that means that frustration comes out in different ways. And the AI just is capable of tolerating all of that with a plum. And so I think there’s really an element there that solves that kind of problem for you. I think there’s other pieces here that just quickly I wanted to run through, which are that One, it allows for Just prolonged practice. And this is an ongoing problem. And again, just another soapbox item. We keep hearing about this research that engaging in physical things is somehow better than engaging in digital or virtual things. Your brain does not know the difference. 

So we need to stop this. It is the amount of energy that you expend on a topic, the amount you dwell on a topic… And that you are challenged on that topic that causes the brain to grow. It is just like working out. When you stress a muscle, right, that muscle literally has micro tears. If you don’t go too much and too hard on the muscle, right, the muscle basically gets broken up. And those little micro tears then force it to reinforce itself. 

And then it grows. Muscle grows through basically a hormetic process where we apply stress to the tissue, and then the tissue responds by getting stronger over time. The brain does the same exact thing. By stressing out the brain in these healthy ways, we’re forcing it to learn about these things. 

So the, Best example I have of this is that when you’re learning something, you’re much more likely to get something right more often. Once you’ve gotten it wrong. 

So it’s this weird counterintuitive perspective where getting things wrong is like, when I’m being assessed, if it says, you know what, this French verb that you’re trying to learn, you conjugated it incorrectly and the proper conjugation was X, Y, and Z, then you go back and you say, okay, yeah. And so the next time you’re tested on this with spaced repetition or otherwise, you’re going to go ahead and now know that thing better because you were wrong before. 

So We need to remember that practice is really important here. Then the other piece is application, right? We need to be able to take that practice content and now take it out of passive consumption and into application. 

So that application could be going back to the learning perspective. We need to comprehensible input. The whole idea here is that you need to be able to take in content in different formats in that language. And so you’re capable of ingesting that language and reading by watching and listening. 

So that all of your faculties start to get used to the language, then you need to then be able to express the language. You need to speak the language in real time. And again, these are all areas where AI is actually really proficient at being able to do so. I think we have a lot of problems in learning. And I think that AI challenges us in really good ways to be better learners. Again, this is just my idea that came from my Substack articles about these kind of four personas. But let’s talk about the AI as a personalized tutor. In what areas are you currently utilizing AI to basically help you be better in that kind of Socratic learning space. 

So let’s just stay in the Socratic learning space and talk about some of the things where you might be utilizing it for those purposes.

Francis | 13:57

I think there’s a difference between Using AI for knowledge retrieval. And using AI to develop personal skills. I think both of them are important and there are different kinds of learning, but AI is a tremendous boost in both areas. I’m going to assume that you’re talking more about the second area though, which is How do I become better at a particular skill, not just retrieving knowledge that AI could give me or could teach me or could pull up for me, but how do I become better at, for example, conjugating a verb, for example, is a personal skill. It’s not just a matter of retrieving It’s something you’ve heard before. 

So I’ve been using AI in particular to improve my writing and to ask questions about how to structure an article, how to… Think about the I learned a whole language from one content producer, which is all about reveals. His name is something Callaway, but he’s on YouTube. And he talked about Writing in a way that reveals different ahas, so to speak, throughout an article. 

So I took his… Which I thought was pretty handy, uploaded it into AI and it was Claude is now using that. Or I’m using it. Or both ways to guide the way that I structure myself the article So over time, It’s learned how to teaching is a strong word, but let’s call it guide how to guide me in formulating an article given the inputs that I’ll give it. And it’s using the Callaway approach. 

So who’s learning and who’s teaching? I’m not exactly sure because I set this up. I asked it. I trained it. With somebody else’s idea. Every two weeks when I write a new article, My idea, which I formatted in a particular way, it takes about six or seven steps to go through all the different sort of the points of the process. 

So I’ve, Learn this new process, but I believe that I taught myself how to do it. But by asking the AI, to do these different steps and to ingest the original idea and to use my content There’s this give and take. Going on between us where I’m asking it for an answer and it’s giving me back an answer. I’m saying, no, that’s not what I wanted. Here’s what I thought of instead. I’ll take it in this direction. No, I don’t like that example. And this Socratic give and take has vastly improved not just the quality of my writing, but having more fun. I’ll actually look forward to this now. In old days, I used to sit down and just bang out a draft and maybe do an outline, but I would just bang out a draft. Now, I am… It says things that are not true. I wouldn’t I was trying to explain this to my wife, Claude is. Remembering things that I forgot. And because it remembers the things that I forgot, I don’t mean like forgot like 10 years ago. 

I mean, forgot like I did a long interview with somebody or I have an idea for an article. I put it together in a long form. Outland? Or a transcript. And when I asked AI to, here’s the idea I’d like to pull from this content in order to, what are the good ideas that you can find from this? It does a better job than I could of retrieving points Then I could on my own. And it’s not because they’re not my ideas. They are all my ideas. They’re all things I wrote down or things that I was a part of creating. It’s that. I just don’t have access to everything all at once. 

So, Am I becoming a better writer? Absolutely. Am I having more fun? Definitely. Not quite. But it seems as if it is because I have access to more of my own knowledge than I ever did before. 

So talk about Socratic.

Augusto | 18:10

But it’s interesting that you said that because as you were saying that, it made me think the same thing. Even sometimes… You start working on a new idea or start working on a way to this new idea and then suddenly or Gemini or Chachapiti tells you, hey, wait a minute, this makes no sense to what you were saying about it. Keep you… A lot more closer into the track of what you do than your own self. And that, without any doubt, has an incredible power because you are really able to work into how you make this thing happen.

Francis | 18:53

– Yeah, and it far surpasses what it would be like to work with even, I think, even the best of writing coaches. Can’t. Replicate the space that’s being created. Between me and Claude or me and Gemini, as you said, occasionally chat GPT. There’s this, it’s not a, better version of our 18 quads. This is A brand new experience. That is distinct from anything that existed before. Really, having used it, there’s just no way to go back to… The way I used to do things. I think the power is in this Socratic makes me think of it being traditionally Socratic way, which is that you ask, you have this Kind of questioning between yourself and the coach. The word Socratic doesn’t seem like a good fit. For what I’m experiencing in the sense that It’s beyond, it is Socratic in part, but it’s way beyond Socratic by the time you take all the pieces. And put them together. 

So there maybe needs to be a new phrase that describes this kind of ongoing give and take. The space that I talked about before.

Raymond | 20:09

What do you believe is that word? I’m.

Francis | 20:10

Curious. I have no language. I’m barely able to articulate it, but I know it because I’m experiencing it. I spent all day yesterday Boom. Part of the day before writing, I wrote a proposal, I wrote an article. And I put the final touches on a presentation and I used AI for all of them. All the way through my, I used my wife 

So she was a collaborator in a couple of those pieces. But I just don’t have a word to put the journey that it is that I’ve been on or the experience or the kind of learning I’ve picked up or the before and the after, let’s just call it that because I don’t even know if it’s Can’t even call it learning in the traditional sense. I just know that there was a before… And there’s a dramatically different after. And that what happened in between It’s hard to call it learning because learning takes me back to something traditional. This is some other kind of. AI machine infused re-jiggered experience where everything has gotten That’s it. Almost every dimension has improved. To wrap all of it up and call it learning. Is I don’t know. I don’t know if I pick up on, if I learn how to prompt an L, And it gives me better answers. Is that my learning? It’s learning. Or is it just some journey in which improvement happened? And we used to call it learning because that’s what it was. No, it’s something… Because it’s automation and automation isn’t learning, correct? But In this instance, automation feels like it’s Somehow coupled with learning to produce a better and different kind of output.

Raymond | 22:05

I think it’s just basically, in a lot of ways, it’s augmenting your approach to things. And so it’s AI augmentation. But there is learning that’s happening in some part of it. And it’s important perhaps not to conflate when AI is supporting you in production of things versus production. Versus the idea of you really, truly attempting to learn something. And it’s funny, I was watching or yeah, I was watching something that came across my social feed. The other day, and I think it was Nir Eyal, who’s talking about children and what we need to teach them today and saying that we shouldn’t teach them things that AI is good at. And I said, I respectfully disagree. There’s no way that you can tell me that we shouldn’t teach children to memorize things with full precision. Argument that he was saying, obviously they need to learn the basics. I don’t know where basics begins and ends for him, but for me, there is no end. We need to teach children how to think. That is the primary purpose of going through any education system. And that requires, I mean, it does not, it’s not an option. It requires them to learn how to learn and it requires them to learn how to basically commit knowledge to memory. And that means consumption. That means comprehension. That means retention. And that means basically being able to then output, basically recall that information on the flip side. And if we just… Delegate all of that to technology, that becomes a real problem for human endeavors. That becomes a real problem for humanity generally. And so I pushed back because I just really fundamentally think that’s the wrong approach to us. Just thinking about the world, there is a value to me when I can summon Shakespeare’s sonnets. And that may not be Today, it may not be tomorrow, it may be 15 years from now, but being able to connect the dots between disparate. Domains is where innovation lies. And so the whole idea of, well, why teach a kid today about the humanities or about the arts or about mechanical engineering when the AI just knows all of those things? And I say, that’s because you never know where those confluence of topics will come together and create innovation. There is no reason why in an abundant society and culture on this planet where we have all the resources to educate all of the people that we wouldn’t do it. It just it creates more meaning when we give people knowledge and withhold that in some capacity to say, like, you’re going to work around the AI just sounds to me so absurd on its face.

Francis | 25:00

I agree and I disagree. Not disagree, but I want to refashion what you said a bit because learning is more than knowledge retrieval. Learning is personal capacity and the personal capacity to retrieve knowledge you don’t have. Is huge part of what we’re learning to do today. Even may. Learning to retrieve content of an interview that I was in. But can’t retrieve on my own because my mind kind of capture all the nuances of a I did a 301 workshop on last week and I remember 2%, 5%, babe. But I know, but I’m… Relaxed because I know how to Access. All the nooks and crannies of the three-hour conversation, not just in terms of what was said, but I know how to… Get at the nuances of the conversation and then use it to take myself to the next level. 

So I would call that learning as a skill as opposed to learning as knowledge retrieval. And then somewhere in between there, there’s learning to retrieve knowledge you don’t have. I think the three of them create this kind of, like there’s three separate things going on there. And I think they all need to be Bastard. And fed and improved. And they all have their place. And I don’t think there’s any, I don’t agree with what Nouriel said about that we need to just delegate. No. It’s not working out that way. Because if you don’t know to There’s certain knowledge that if you don’t know that it exists, you can’t retrieve it. You won’t know to ask about it. Toy. Is wrong in that sense. It’s not nice and linear and black and white. There is many shades of grey at the moment. And we need all of those three kinds of learning to make progress.

Raymond | 26:53

On your point about asking the right questions, one doesn’t know how to think correctly, even with all the knowledge in the world. And so I liken… AI to basically having a million interns. And all of the analogy that comes along with the concept of having a million interns in your pocket And so the idea here is that they lack life experience. And so even if we were at some point embedded with a chip, right, you just plug AI directly into the human brain. Now it has this high capacity for calculation and high capacity for holding large bodies of information. Does that actually make me better if I don’t know the right things to think, the right thoughts to summon? 

Like, it takes away some of your humanity. If you relegate all of that to the AI is my personal philosophy. Maybe you have a different personal philosophy on that and you want to be embedded with a chip and stop thinking and just have it make all the decisions and defer your life to it. I don’t think most people would choose that if they had agency over that subject. And I think that is a form of subjugation in a lot of ways and dangerous in others. But The idea here is that You challenge in your life the learning as a mechanism for being able to grow as an individual and grow. I think it’s grow or suffer worse in life. Why live a life where you are not continuing to grow in some capacity? And this actually plays into biological success. Whether you have children or not, you want to be able to Pass on the knowledge you were learning, the betterment of your life to other people in some way, shape or form. And if you don’t get better, then you don’t have something to contribute to the next generation. And so you become less useful in that way. Surviving is not thriving. And I think surviving is using AI to take over things that you otherwise would have agency over. And that’s just, I guess that’s just my opinion on it.

Augusto | 29:03

It is interesting what you both said, because yes, AI has the risk to have the knowledge for the knowledge or to bring us to knowledge that may not mean anything, but But I think That is the same thing as a library had at some point. You can sit in the library, get into the library and find books and pile books, okay, and get knowledge to be knowledge. And we know people who do that. And that’s where I disagree with that person you were mentioning early. No. I will work. After a certain level. You need to at least have the basics and the basics will depend person to person what you do and it will change and it will evolve. But if you don’t have the criteria To understand that, then is going to be knowledge to get knowledge. And that’s completely useless.

Raymond | 29:56

I wanted to talk a little bit about some of the ways in which I think AI can be really useful in learning One is I like to, again, just thinking about More AI, my AI learning partner, I always want to have a kind of Socratic mindset. Mode on. And that allows me to, in essence, have it do a couple of baseline items. One is to consistently challenge my assumptions. And so I don’t want it to be sycophantic when I’m learning. I want it to take whatever I am trying to learn about and challenge me on how what I currently know applies to it. And so I will frequently ask it to challenge interrogate me. And that just means interview me, but I use the term interrogate Because I want it to be hard-nosed. I want it to be like an investigative journalist. I want it to be biting with regard to its questions of me. And that is about what it is that I know that might be connected to this particular topic or might be in different domains that overlap in that specific subject. 

So I really enjoy being able to say that to the AI and have that take place. In either a role-playing scenario or a scenario-based learning environment. 

So… Having it simulate different environments where I need to use that particular knowledge, and that can be really helpful as well. 

And then finally, I like to have it create context. Personalized curricula. 

So that… I can say to the tool, identify my knowledge gaps and then generate the next module, the next learning session, the next piece in the syllabus to teach me that thing so that I can get immediate and targeted feedback along the way as I am trying to learn. And again, this is a continual kind of question and answer style approach. Where once I’ve gotten it wrong, again, as I said before, getting things wrong is really good for learning because once you’ve gotten it wrong, then the AI will correct you and give you the right answer back after it’s pushed you on trying to get to that right answer. And that means you’re going to learn it a little bit better. And so that’s been the core of what I’ve been learning and using it for. I’ve been using it for language learning. There’s two different. Philosophies there that kind of mirror each other. One’s comprehensible input, the other one’s comprehensible output. 

So the idea of just ingesting material in a foreign language, the other is being able to just immediately apply and speak in that particular language or write in that particular language. So the idea here is that AI can be really a helpful partner in being able to engage you in that process. 

So moving right along, let’s talk a little bit about skill acquisition and how do we shift from acquiring knowledge to mastery of a subject and how can AI really help us do that? So how do we ensure that the learned skill is both durable and transferable? And what boundaries and ethical considerations should we think about in terms of the AI learning partner, especially concerning over-reliance on the AI and maintaining a human-to-human collaboration perspective? Do you all have thoughts in terms of how you treat the AI in this particular capacity?

Francis | 33:25

So a little bit of a story. I was developing a site to share stories from YouTube on YouTube. Different parts of strategy, different strategy stories, like what happened to IBM, what happened to Motorola, these things. Kodak versus Fuji and during the process of putting together this site, I realized from listening to a YouTube video. That I was not in the education business, I was in the entertainment business. That came by accident, I hate to say it. Because I had called the website Strat Cinema and Shameless plug, stratcinema.org is now live up and running. But at some point I said, hold on, if I’m calling it Strat Cinema, then why do I think it’s going to be for education? Because you don’t go to a cinema for education. 

So what the heck am I doing? So having been prompted by these YouTube videos that said that, The idea was that you either have an entertainment channel on YouTube or you have an education channel, but you don’t have both. That idea led me to think that I wasn’t thinking about for example, conference learning in the right way. And it gave the example of a World Fair, So this is before our time, I believe. One of the world’s fairs actually came here to Jamaica in 1896 or something like that. But the whole idea was that you’d go to this fair and see all these amazing things from all over the world is amazing. And the only way to have access to them was to go to the World’s Fair because there was no internet, there was no TV, there was not even any radio. You had to go to the World’s Fair. But the point of going to the World’s Fair was to gain exposure In a fun way. A little bit like what happens when you go to Epcot. Or any interactive museum, where actually most museums that I’ve been to in the US have interactive components. But the point is that you’re not going there to learn something. You’re going there to gain exposure. 

So you don’t have a particular objective in mind. You’re going to enter into a fun experience where you’re exposed to lots of potentially Fun and interesting, I hate to say pieces of information, but let’s just call it experiences. 

So I took the idea and I said, you know what, that’s what Strat Cinema really should be. It should be a passive learning experience. But the point is that You need to have this passive experience in order to feed what I would call hard learning. 

So hard learning is when you need to pick up a skill. You need to be able to do something that you couldn’t do before, like vibe coding. But the way to get the first exposure is to have it be passive, which means that you need lots of input So preferably entertaining, educational, thought-provoking. 

Like you’re walking through a museum and you’re like, look at this. Look at that. Look at that. In that mode, you’re not trying to pick up a hard skill. You’re in this other kind of open-ended kind of learning. However, once you do exit that stage and you’ve picked a topic. To learn on and the game changes, but you’ve got to be conscious that you’re no longer visiting and passive, you’re now entering into an active mode. 

So it’s not like being in a conference. It’s more like after the conference, you go find the speaker and you sign up for their six-week course. Or you hire them as a coach. Or you buy their book. And you don’t read the book for entertainment. You read the book. All of us have done this. You read the book three, four, five times. And you underline all of the bits. And you start to think that, you see the counter arguments. 

You see that you should, they mention this part. That doesn’t quite hold up. You go do deeper research. You realize that the research supports it. 

So you’re doing all these things to enter into this What I call hard learning. Hard learning means that at the end of the activity, you’re engaging in some skill that you didn’t have before. Or you’re retrieving knowledge that you didn’t, like conjugating a verb that you didn’t possess before. And it takes serious effort to engage in that. 

So I think there’s both this passive stage to go through that can be fun, And then there is this Not so fun. Very active very focused stage and AI is amazing for both. It’s just different. You’re just using it. Very differently and you’re looking to have different kinds of experiences And you don’t have this experience of being gay guilty, for example, of not learning AI, because right now there’s a lot of AI that’s being taught. And there’s no way to absorb all of it because there’s too much of it. In the sense that there’s too many new things happening. 

So we have to be in this passive stage for extended periods of time where we’re just basically visiting different capabilities, concepts, apps, LLMs, software as a service, whatever it is where we have to sample in a way that we never had to sample before. And then we have to choose, okay, I’m going to focus on this skill. 

And then we apply all of our efforts to learning the single skill. The difficult skill. We need to know that these two modes are available to us and we need to know when to change gears from one to the other, either to flip into the one or flip back to the other so that we can just retain a sense of sanity and not feel guilty that I got all this AI is happening and I’m not learning how to Fill in the blank with some difficult skill that you haven’t learned yet. And you cannot learn because you just don’t have enough lifetime to pick all of the skills that are now emerging. You cannot learn all the skills that are emerging in this lifetime. They have to be Selective. And you have to choose and you have to decide. I think this is a huge challenge, Ray. Selecting the deciding and the choosing. Are a major challenge for us in this new AI driven era. It’s the curation of our It’s a curation of our learning experience. It’s gone from just reading a book and then applying what you read in a book. Now we have a fire hose. Everywhere we look and curating our Progress. Is Huge thing. This is almost like hiring yourself as a coach to manage the next five years of your learning.

Raymond | 39:55

Wow. I agree with you in the premise of things generally. I think some of the specifics I probably would use different language. I don’t think passive is the way in which I would categorize that first stage. I think it’s consumption, but it is active consumption. If you’re not engaged in the material, I think, for example, I watch a lot of documentary. Films. And I’m actively consuming that content and I’m engaged because I’m challenged. 

So again, going back to a little bit of neuroscience here, neuroscience light. But the idea here is that you will not learn something unless you are challenged by that material, which is why entertaining things are usually more challenging. Engaging, more active consumption. I don’t necessarily know that you always need to be in a entertained mode, but when things are funny or when something is in that entertainment space, you’re much more likely to, because a joke is basically hook and then line, right? And so the idea is that you are consistently engaged in this kind of like peekaboo. It’s just basically an adult version of peekaboo, right? And so we just have object permanence now. And so we go through this constant like peekaboo stage. And so the active consumption needs to be there. You need to be challenged by that thing, right? 

So if you want someone to really get engaged in something, you then you bring up something that is shocking for them to be able to retain their attention in the consumption process. So you say, you know what? I love bacon and pineapple on my pizzas. And all of a sudden, someone’s like, well, I don’t like pineapple on my pizzas. And if you can now relate that analogy to whatever you’re teaching. You’ve got them, right? You’ve sucked them in. You’ve made them. I always think about when I’m teaching, are they leaning forward in their chair or are they leaning back, right? And if they lean back in their chair, they may be contemplative. But most of the time, I want them to be in that active state where they’re ready to challenge me, right? I want them to say, you know what? I don’t agree with everything he says, or this point he just made really connects to this other thing that I was thinking about. And that really makes a lot of sense. And to go back to my documentary film consumption, I like that because I’m learning because I’m challenging it against what I currently know. What’s my current doctrine about this subject? And what is it? Telling me new about that, whether that’s a person or a thing or some kind of larger philosophical topic. And so there is that passive consumption side, but I think you’re right that you then You graduate beyond that if you want mastery. You need to graduate beyond that because you need to make sure that you are applying those skills in some way, shape, or form. And this is where, again, AI can just be really useful. I think it could be useful on the consumption side because it could be acting as… 

I mean, I have for 25, 30 years been in constant conversation with every book I read, right? Constant conversation with the author because I’m a heavy annotator as I’m reading. I am in dialogue with that author. And I think there’s a lost art in annotation in the margins of books. And maybe someday I’ll write a book on that. But the idea is that I really fundamentally believe that if you’re not in that active consumption process, And then taking it beyond that to an active engagement process in the learning, and you’re not going to learn as well. And so I think AI can help in almost every one of those capacities. Augusto, and then we’ll do final takeaways and close out.

Augusto | 43:30

You made a very interesting point. And like you, I annotate. Not as extensive as you do, because I’ve been limited for years into… Highlighting on the Kindle. Okay. I went to digital books. Many years back, it was… It took them a while to give you the quick note that you can add it. 

So there are many books that is just highlighted. I remember when I found a tool that could go into the Amazon libraries and extract all those. Now you can extract all those and dump it into the AI, that it makes something very interesting because as you add new highlights, even if you add a little note, it really helps you to connect all that knowledge in ways that I bet. I don’t know if I have the skills or the time to really go back to that first productivity book, okay, that I read. That book that I read, and it was 200 pages, and I highlight one paragraph. And I have some of those, okay? But you don’t necessarily remember that paragraph fully, okay? And since I said at the beginning, Some things you will acquire but don’t know how to use it. And I quote and I will quote forever that Alvin Toffer line, okay? “The alphabets of the 21st century are the ones who cannot unlearn and learn again.” Okay? I learned that when I was a kid, okay? And it was interesting. Did I understood it? No. It took me years to get to the understanding of that. But that one in particular, as knowledge, for whatever reason, I hold. The problem is, as Francis was saying, we do not hold to all those pieces. We may found it interesting, we may found it intriguing, but we don’t hold it. One of the great things of AI is that it will hold it for you and it will bring it back with a little bit of luck, when appropriate.

Raymond | 45:33

As we take our leave of this episode, let’s talk about what our final takeaways are. If someone is trying to apply AI to being this learning partner, this learning coach role. What are some things that they can do to get started.

Francis | 45:51

I would encourage folks, someone who’s listening to pursue their interests using the new tools that exist. The I’m still shocked and surprised by AI and LLMs even books that I’ve already read. Uploaded into Notebook LM or doing a search in cloud or an open chat, I’m amazed at the ways in which things I already think I know can be refreshed. Because I have a learning partner. 

So I would encourage folks that whatever your, it doesn’t really matter what it is. The point is, it’s a little bit like going into Epcot and you know they have an exhibit. I don’t know what’s in Epcot now, but let’s say they have an exhibit on dinosaurs and you love dinosaurs. It’s to go and use these tools to go pursue any area of interest that you have. And just keep relentlessly pursuing it because even the stuff you’ve already encountered, stuff that you already think you know, Stuff that you may not have even taught other people. Is no subject to More fun. And more deliberate learning and more specific kinds of skills and they’re now available that the landscape around anything we thought we already know or anything we are now interested in It’s changing rapidly and it’s becoming way more engaging and way more fun. 

So I just, I think it’s just have a lot of fun and just pursue and just see where it takes you. And It’s got to be something deeper than that. When you started. Gotta be. Because that’s just the way the world is going right.

Augusto | 47:32

Now. I will say a few things. If you… If you trust, if you’re comfortable with AI, Just go pick a topic. If you are not, then pick a topic that you feel confident. Peter, pick a topic that you feel You are. You are very confident, you’re comfortable and learn from that. Because what you’re going to find, at least in my experience, is that you are going to get into a new level or a new deep level into this process with that. And that will give you the confidence to then go and run with a new topic.

Raymond | 48:08

So I have a few thoughts here in terms of Some takeaways. One is if you want to ensure high quality synthesis from your AI assistant, you need to really provide them with explicit instructions on the format and structure of the desired output. 

So if you want to be able to have that kind of Socratic structure, Skill Acquisition. And master a topic, then make sure that you are explicitly telling it, I want to do this in a Socratic learning style. 

So make sure that you are not sycophantic and Make sure that you are challenging me at each stage of the learning process, giving me fast feedback and. Quizzing me. Randomly. And just those kinds of parameters can really be helpful. Always insist on citations if you are doing this in text. 

So you can say like, I want you to make sure that you cite everything that you are sourcing for this particular knowledge base, because garbage in and garbage out. So if the AI is giving you wrong answers about a topic, then you being wrong is not helpful. It’s just reinforcing the wrong knowledge. And it’s very human because humans teach the wrong things many times. 

And then I’ll close by saying that I think there is a form of deliberate practice where you can go ahead and use that deliberate practice perspective so that you can move forward in the learning process. And I’ll put a link to what deliberate practice is and the tenets of it. In the show notes. 

So folks can go check out what deliberate practice is. But you can, in essence, tell the AI. I want to do this through deliberate practice and set up the syllabus for such and engage in a Socratic method in the process. And voila, you’ve given it the key parameters, the key boundaries for being able to engage the AI in this really practical way. And so with that, we’re going to close up the episode now and we’ll continue this AI-powered professional series with talking about AI as a coach in the next episode. But thank you, gentlemen, for this listening. Fun topic on learning with AI.