SPEAKER_00: 1:31

Welcome back to Inspire AI, the podcast where we explore what happens when technology stops being just a tool and starts becoming a collaborator. I'm your host, Jason McGinthy. Today's episode sits at the intersection of human potential and agentic AI, and it's one I've personally been deep into lately. Experimenting, breaking things, rebuilding them, and rethinking what coding actually means in an AI first world. And that's exactly why today's guest is such a strong fit. Today we're joined by Godwin Josh, someone who's spent over 20 years working with highly intellectual teams and individuals helping them solve genuinely hard problems. What stands out about Godwin isn't his technical depth, it's his pattern recognition around people. He's seeing this repeatedly. The world is full of intelligent individuals. But intelligence alone doesn't translate into impact. What does translate is mentorship, problem framing, and exposure to real-world constraints. And this becomes incredibly relevant when we talk about AI today. Because AI isn't just changing what we do, it's changing how people learn, how they reason, and how they become effective much faster. Godwin works closely with universities, professors, and students through initiatives like Spark, where students are taught theory in isolation, but are given real problems to solve with the explicit goal of becoming work-ready and recruitable. That land, learning through real systems, is exactly how AI coding is evolving. What we're seeing now is a shift away from coding as a syntax-heavy manual activity toward coding, as an act of reasoning, orchestration, and intent. I've personally been spending a lot of time experimenting with agentic coding tools, watching how AI compresses the learning curve in ways that simply weren't possible before. Tools like Winstur from Codium, Claude Code from Anthropic, an entire ecosystem emerging around MCP servers. Where agents don't just generate code, but coordinate tools, context, and execution. This is where AI stops being autocomplete and starts behaving more like a junior engineer, a systems thinker, or even a collaborator. And that raises some important questions. We'll unpack today. What does AI coding actually mean now? How do agency systems change the way developers think, not just work? Why are tools like Windsorf and Claude Code less about speed, more about learning and acceleration? And how should educators, leaders, and engineers rethink skill development when AI can fast-forward confidence? This is where Godwin's work becomes especially relevant. Because when AI removes the friction of syntax and repetition, what's left is intellect, judgment, problem framing. And those are human skills sharpened through mentorship and real-world application. Godwin's book, The New Mind, explores this shift from the ground up, not teaching AI as a buzzword, but as a new cognitive model. How people learn, think, solve problems in an AI-augmented world. So today's conversation isn't just about tools. It's about how agentic AI changes learning curves. Why coding is becoming more conceptual and less mechanical, how students, engineers, and leaders can become effective faster. And what work ready truly means in the age of intelligent systems. Godwin, welcome to Inspire AI. Thank you for being on the show today.

SPEAKER_01: 5:26

Thank you. It's my pleasure.

SPEAKER_00: 5:28

How about start by telling the audience a little bit about your background, what brings you here? And also where are you located?

SPEAKER_01: 5:37

I'm located in Chennai in India, city by the shore in the southern part of India. And I'm very happy to be speaking on AI because and technology, because that's what I've been into. Over I've been into tech for the last three decades. Over three decades since my childhood. I've always been into computers since the mid-1990s. That was the first time I used a computer. My father had a video studio back then, and I had access to the 286 and 386 computers. Started with making animations in DOS. We used to have a pro, you know, a software called DOS Animator Pro. So I used to do small gigs for my dad when he was running his business. I used to go into his office and just do small things. That's how I got started with it. I made my first website in 1998, and uh that's how I got into it. I I I just loved computers. I got into programming eventually, tried all sorts of tools. You can call me a fiddler. I used to always fiddle with computers all the time. So, and then eventually I started a team. I founded a company. Initially, it started as an IT department in my father's business, and then I moved into a separate company. I had my office, and along the way, we've been developing a lot of products, hardware products, software products. One of the products was an internet generator that could combine multiple internet connections from multiple SIM cards and provide a consistent broadband connection on the move. So if you're traveling, you're going on a vehicle, you can have this hardware box which will give you consistent internet on the move. So that was one of the products. Then we made video encoders, decoders, streaming servers, a streaming server called LiveBox. It's used by a lot of clients all over India. Initially, a lot of cable TV channel operators were using it. And now people use it for transmitting and receiving uh video across distances. They use the bonding unit. So these are some of the products. But as time went by, I eventually saw with the flow of technology into AI. Once I saw AI, I just loved it. I'm so hooked into it. And now I'm so focused on building teams on AI, not only building teams, but also helping them with research and development. So that's that's what I'm into.

SPEAKER_00: 8:12

Okay. And you wrote a book recently.

SPEAKER_01: 8:16

Yes, I've authored a book called The New Mind. Ever since I encountered AI, I wanted also to write a book because I that that was a huge eye-opener on how it came out. Because even before I you started using AI, I always heard artificial intelligence, but I never really believed in it. Of course, I knew what was going on. I knew about the image classification and and and the way it recognizes patterns. I knew it, but I didn't really believe that it was gonna take off this big until the moment I actually saw Chat GPT, tried it with my own with my own hands. I'm kind of like doubting Thomas, you know. I didn't believe it until I felt it.

SPEAKER_00: 8:58

Yeah, no, I think that it took the world by surprise, Chat GPT in 20 the late 2022.

SPEAKER_01: 9:05

Yeah.

SPEAKER_00: 9:06

And when when you realize that you can just type in natural language into a computer and then it generates content for you, it woke a lot of people up, for sure. So you you had known about convolutional neural networks, RNNs, that sort of thing, before Chat GPT, before the rise of Transformers, and you expressed a much larger interest when you saw it in action. Tell me a little bit about your teams. Um, tell me a little bit about what you empower them to do.

SPEAKER_01: 9:45

Okay. So before Chat GPT came, I had absolutely no clue about Transformers. I was not into that at all. I only knew an overview about AI. It is only after I saw Chat GPT in January. I actually saw it first in January 2023, all that came out in November. And just after that, my you know, really like you said, I also started just woke up, like, what's happening? And then I started learning a lot about it. Who created this, how it's evolved along history? Because I've been there all through that time, but I was never involved with what's happening into AI. So I wanted to really find out from where this came from, who were those people who made this thing, where were they working on it? And those are the kind of things, and also about Transformers and and and uh CNNs and all that. I I wrote about it on my book. But with regards to the teams, I generally, you know, one of the things I do is I mentor young people and I help them. One thing is, you know, forming teams for my company because I need people, and most of the people in my company are mentored by me. They joined a company about 15 years back, and then I coach them, train them, and today they are experienced people. I feel that very comfortable because they've been mentored all the way through into a company. So, in addition to that, we also partner with universities where we train, you know, mentor the students in those universities, form, make them form teams among them themselves. So let's say there's a group of people who are who are specializing in physics, there's another group of people who are specializing in data. A lot of times these people don't come together because they're in their own groups. So we bring them together to form different kinds of teams, and then we take a problem statement for these teams. So, okay, the guy who learned physics and the guy who's good at data science, who's good at programming, so they bring their knowledge together. So maybe this person wants to, they want to work on something related to space tech. So they need somebody who knows physics, who knows about gravity, who knows about motion, who knows about all these things. So we bring these people together, we take a problem statement, we allow them to brainstorm, and then we allow them to accomplish, you know, help them hit the goal. So that's one of the things we do. And that way we're building a lot of teams with universities in India. Currently, we're working with about five universities. We help the students uh innovate and make things, do projects even while they are studying in their university. So some of those students we pick them up from the second year of their university, and we start mentoring them, teaching them how they need to think. Because I think in the AI age, it's not about just knowledge. Knowledge is just one part, but there's more to it than the knowledge. How do you think? It needs a change in mindset. So that's that's one of the things we mentor them through from their second year of university all to the final year to make them ready for work. So they're not a student when they come out, they come out as people who can think and solve problems.

SPEAKER_00: 13:12

Awesome. Can you can you give me an example of a problem that these individuals with specialties come together to work together on? It sounds like they they know what they're getting into when they get there. They maybe know a little bit about themselves, the the group that they're gonna be working with. How does all that come together and what what problems do they solve? Can you give me an example?

SPEAKER_01: 13:33

Yeah, it it could be any kind of problem. It could be even making a software product, it could be maybe finding out to f you know, find finding out new forms of proteins, like so it could be any kind of a problem, but the most common ones are software problems using software, you know, making agents and things like that is one of the things. But that is just one of the things. The other things is even because with AI, you have we have so many tools that people are not aware of. For example, Nvidia itself has so many tools. Earth 2 is one tool, for example. There's a virtual earth, you can simulate, you can simulate the world in a virtual way. I mean, people don't know about these things, so we expose them to those tools, we help them solve problems on that. So it could be even drug discovery. So some students choose that, they want to discover a new drug. We help them along the pathway so they can get there and find these new drugs. So at least attempt to find. We're not saying that they are going to completely find and invent it, but at least we take them through that route so that when they finish their university, they are ready to get employed in a company that is working on the same field. They know the stuff and they can really help with that process.

SPEAKER_00: 14:52

Okay. So you it kind of sounds to me like it's a mini internship of sorts where you're connecting people and real-world problems so that they can build hands-on experience outside of their university classroom instruction. Is that right?

SPEAKER_01: 15:09

Yeah, it is kind of an internship, but but the thing is, this was not possible in the days when AI was not available, not easily possible. So now it is, you know, people have access to huge amount of knowledge at the tip of their fingers. They go to chat GPD and ask, hey, how does this work? How does that work? How does this work? So this guy, maybe he's already always learned theoretical physics, he does, he hasn't worked practically. And these guys go to AI and they start they the amount that they can accomplish is huge. What a student never had access to before. Now they can do it, and that's what makes it possible. And that's what I mean, that this is only possible after the AI age. The pre-AI era, I think this would have been really challenging, really challenging to do.

SPEAKER_00: 15:59

Yeah. Well, so let's take an example then inside software engineering. Let's say you have a team that wants to get together and solve a software problem. What kind of tools would you say that they should be starting with? And then maybe they're maybe they've already had hands-on tools, and then what's the next level for them?

SPEAKER_01: 16:17

Yeah, the one thing that we think is common for everyone is Linux. I think most tools need Linux, and one thing that's common for most of these people is the Linux operating system. So we specify that they have to learn Linux, they have to learn Python. That is also common. At least know it's not that they have to be very deep into it, but at least know a little bit about it. So these are some of the basic things we require them to learn because most of these applications or problems they solve require these. Not all of them require it, but most of them do. So that's part of what they need to learn. That that comes as a basis for most of it. But other than that, there are other tools. I'm just not getting the name. There's a tool from Google that helps you find new protein structures. So, I mean, that's one of the options available. There are multiple tools like that. There are different tools for drug discoveries, for discovering new kinds of chemicals and things like that. So we give them these tools and they work with those tools and they accomplish that particular goal, whatever they're trying to accomplish.

SPEAKER_00: 17:29

Okay. Do you do you ever see them working with coding tools?

SPEAKER_01: 17:34

Of course, agency coding is now mandatory, it is part of every work, and that is also so fundamental today. That is fundamental not only for the students, but also for our developers, for any anyone who listened to development. We think agent is agency tools is just the fundamentals, part of the basic.

SPEAKER_00: 17:55

So if to give you an can I walk back to the first observation we just made about the tools that you say are prerequisites, Linux and Python, would you say that those are still prerequisites to have a baseline knowledge of those things now that you have those agentic coding tools available?

SPEAKER_01: 18:14

Yeah, I still think that knowing the fundamentals where these things came from is essential. And the reason, of course, we can choose any programming language, but the reason I specify Python and I specify Linux is because most of the tools are available in them. I mean, you can use anything, it's it's a freedom of choice, but these are most of the tools available. You you need a package, you just put pip install in Python. Most of the packages for AI, majority are available there. But if you take, let's say, for example, something like Rust, it is available there as well. It's available in C Shop, but you have a larger number available in Python. Likewise, why Linux? Because a lot of these applications, you can't install them like you know, there's no setup.exe that you can just double-click and run and install them on Windows. You need Linux for it. So that's the reason we you know specify that.

SPEAKER_00: 19:13

Uh I I totally understand it. I'm just trying to bring our audience along with us on this journey. So, what AI coding tools or agentic tooling are you using these days? Or even beyond that, like what gets you excited about these things?

SPEAKER_01: 19:30

Yeah, we use pretty much all the tools available, like from Clot Code, VS Code, Copilot, WindSurf, Ker. We I haven't used Cursor so much, I use Windsurf, you know, and a lot of other tools like Tray, ID. Yeah, a lot of these tools we've tried. And but I think it doesn't matter which tool, agentic tool, I mean, agentic tools come in different forms, and one form is the ID for coders. There are other agentic tools that don't specifically code, they do a lot of other things. And but but I think in my perspective, if you just have access to an agentic ID, you can pretty much do anything, you don't need any any other agentic tool. And there's so much to learn that because there's so many concepts, so many things to learn in in these IDs, and now they're getting standardized. Previously, you had different IDs that have a different way to provide custom instructions, a different way to get them work, but now things are changing, you know, like you have these markdown files that are getting standardized. And uh I can say is that today I think it's not it's not worth writing some parts of the code by hand at all. I think people shouldn't even write code by hand. Let's say you want to make a REST API, you have a function, you want to make a REST AI, you know, or maybe you have five functions you want to make REST APIs out of it. Why would somebody go and type that by hand? It doesn't make sense, and AI is not gonna hallucinate for something like that. You just go to AI and tell it, hey, just type this into REST APIs, it's gonna do the job for you. So some of those things is just straightforward, and I think that's the most basic way a person can use AI, at least you know, just giving it a chat GPT. But beyond that, using agents, there's a lot you can do. And yeah, I have a lot, you know, there's a lot I can share about that. For example, now you have agents.markdown file, which gives information about a project. You have a project. What is your project about? You have your once your AI goes to the project, doesn't matter what ID you use. You can use wind surf, cursor, uh, plot code, doesn't matter what. As long as you have an agents.markdown file, once your AI that is using whichever ID looks at your project, it understands the project. It you know, it's like the project explains itself. Hey, this is what I'm all about, and this is what I do, these are my features, this is how the code is organized, these are the folders that exist in the project, it's just gonna explain by itself, and now there's a lot more. You have skill markdown files, so you can specify skills. So, for example, you want, you know, maybe there's a developer who's working specifically on videos and on a video platform or some project like that. So they would have a skill that is so tailored to use FFmpeg. F is a very common tool for you know, just what is it called? FFmpeg.

SPEAKER_00: 22:43

FFmpeg. Okay. I have to look that up.

SPEAKER_01: 22:46

Yeah, FFmpeg is a common tool. So you would have a skill file that just specifies what how to use FFmpeg and uh crop the video, edit the video, do whatever using the video. Once you have the skill file, you just talk to your AA and say, Hey, can you just go and grab that video for me and just crop it into this and also just uh cut the first 30 seconds and save it in a in a folder. Now, when you tell this to your agent, your agent does not have to be wondering, should I use this, should I use that, and doesn't have to go into a lot of thinking mode. It just uses the right skill for the job because these skills are like reusable stuff, prompts. It just uses that skill, gets the job done for you. So that's a skill file. So Now I think what developers have to do is not write code. They have to write plain, simple, plain English prompts. Write skills. How should it be done? Write the agent markdown files. And now there's also this new thing called a soul. That is very interesting. You can even specify the the Did you say soul like S O U L? Yeah, that's right.

SPEAKER_00: 23:57

Okay.

SPEAKER_01: 23:59

You can actually specify the soul.markdown is another markdown file where you can specify the personality of your AI. So when you're talking with your AI, how do you want your AI to think and respond to you? How should I talk? Should it always consider making codes or should it always generally do certain things like examine your files, do testing, or always consider security? So, you know, previously we had all these things as custom instructions for WindServe, custom instructions for cursor. Another way to write custom instructions for VS Code. Now all these things are standardized. These are standards across all these platforms. And you know, first MCP came in, and then all these things started coming one by one. MCP, MCPs are like, see, we've always had REST APIs. Okay. So you had a you, you know, we we could always call REST APIs, make REST APIs, and develop using REST APIs. Then why MCPs? Because these large language models, when they look at the REST APIs, they have to figure out what to use, how to use. There's no standard documentation for the REST APIs. Of course, we had Swagger and all those things, but they're not like so standardized. This is the way REST API should be. If you ever make a REST API, it should always have a description of how it works. It wasn't like that. You could just make an API, doesn't necessarily have to have a documentation, could be named in any way. But with um MCPs, it was more formally standardized. If you ever have something or uh a command or a function, it needs to have a description of what it does. Full description, what it what it returns back. So, you know, these MCPs were more like self-descriptive functions. Hey, this is what I do. I'm I'm I'm this video editing tool, and I do this, and I can do this for you. So it's like it has a whole introduction about it, and they standardize the way these large language models speak to these functions. So they they they use the standard JSON format. That's very similar to how how the old services work, the same thing, but I believe it's the same concept. They just reinvented the concept for AI. That's what is MCP. And Anthropic did it in their own language. They said, this is like a USB port. That's what Anthropic calls it. It's the USB port. You it's like a standard format. You want to plug it to AI, you don't need to say, hey, find that API, find that API, find that API. You just take this MCP, plug it, and AI knows how to use it. So yeah, so so it's great.

SPEAKER_00: 26:44

I'm totally following along. I I want to touch on something though that it sounds to me like you've already transitioned through. And that is the anxious feeling of you know, the transition, right? If you're in a mentorship with somebody, how do you prepare them to thrive when AI systems are handling more of the how now? Especially when they come from a background that is deeply hands-on, and they say, Oh, I have 20 years of coding experience, this AI will never do that for me.

SPEAKER_01: 27:19

Yeah.

SPEAKER_00: 27:20

Right?

SPEAKER_01: 27:20

Yeah, yeah, yeah. Jason, we're actually moving into a new era. I think this is a totally different generation. It's like, you know, like my if my if I speak, if I when I used to speak to my grandparents, they used to say, you know, those days we didn't have, we did not even have electricity in some place. We didn't have a cell phone, we had to go and stand in a queue to to pick up a phone and speak. It was not easy to communicate in those days. We had to walk long distances. Look, when my grandparents looked at their next generation, they had gone far past into the future. So it's like a different generation, they couldn't even comprehend the internet that was so new to them. And I think what we are seeing is an overlap of the next generation with our generation. So I think that's where we are. So when a lot of the experienced people like us look at it, we are like, oh, what is this AI doing? What is it going to do? Basically, what's happening, what I think is that it needs a shift in mindset to understand that hey, the world is changing. We need to think differently. We've always been used to coding, I've been always used to doing this. I'm used to always following these principles or you know, these coding standards. I look at these documents, I follow these 10 steps and do this. But now we have to break our minds out of those 10 steps, and now we have to think differently. Hey, now a lot of that work, AI is doing it. So, what do I have to do next? I need to now ask the right questions. I think asking the right questions is the first thing. Asking questions is so important, and this is something common that uh even in my mentorship, I teach both students and I teach both experienced people. You know, one of the things is asking questions. I know these are not these are not technical things, but these are important. Asking questions, being bold enough to ask those questions because you can ask questions, but you may think, oh, can I ask it? Can you not? But I think now we have to just change our nature, be bold to ask questions. If you don't know something, just ask. And I think a lot of things start from asking questions because previously, if you ask questions to somebody, they say, Oh, don't keep asking me, go and find it your you know by yourself. But now we have an AI that can answer whatever we ask. So asking that and also writing these prompts, that's important. More than writing code, you have to focus on writing these skills, these agents.markdown, these custom instructions based on your experience. Maybe you have a lot of experience in coding. So you you bring all of those things onto the table and you you you write, hey, whenever you write, you tell the agent, whenever you write the code, never do this stuff, never put too much of loops, never put the variables like this, don't create memory leaks for me. So you kind of train and tame the AI to the way you want it to work. Now it's your AI. And remember, it's now not just one the day, present day, each of these agents use a swarm of AIs. So you actually have superpower. Each of these people have superpowers. And I think you just have to recognize hey, today I have a great power, I can do a lot. And one more thing is that we, you know, it's like getting off from a bull cart and getting into a car. So the rate at the pace at which we need to learn has increased, but we also have a different vehicle. We're not, we're, you know, nobody's telling us, hey, ride a bull cart faster than a Ferrari car. No, they're saying get out of the bull cart, get into the Ferrari cart and step on the accelerator. You will reach your destination quickly. So they now we have the right tool so you can learn faster too. For example, you can watch a three-hour YouTube video. You just say, hey, summarize this video for me, what are the top 10 bullet points? You got to answer, you have the tool. So it's really not complicated. You just shift your tool and you can actually accomplish what you want. But it needs a shift in mindset because if you still think that, oh, the car that's just too new for me, I'm just used to the bullet card, then I think that's basically a shift in the way we think.

SPEAKER_00: 31:32

So, Godwin, as as AI agents become teammates, what new mentorship challenges are arising?

SPEAKER_01: 31:40

Like I said, the thing that humans need to change is, you know, until now, if they were thinking that, hey, I need to learn Java, I need to learn C, I need to learn this, I need to learn about web services, I need to learn about full stack. No, that's changing. Now it's better to learn more of I, at least this is my perspective. It's worth learning more of math. It's worth learning more of language and communication, writing prompts than learning all these small, small, small, small, small, small knowledge nuggets. The basics are more stronger as a foundation and they will not change right now. Your math will not change. But if you learn something new that's coming up, maybe even these MCP servers, for example, they'll change. Tomorrow it'll not be the same. But if you learn math and language that will never change, and I think we have to shift our focus there. It's AI is now bringing giving that power back to us. You know, no need to be focused on syntax-based work, but on intellectual work. And I think now intelligence is more uh valuable than before. You know, in it's now what matters is only intelligence. And like I said, what needs to change the way we ask questions, the way we make decisions. So if you are, you know, some people, for example, they're used to like asking, what do I know next? What do I know next after this? Oh, yeah, I and then what do I know next? So, you know, so those are some things that I think people need to now change. Decision making is another another key thing. So, one thing what I was saying is asking questions, another thing is ability to make decisions. Hey, before I do something, let's let's say, let's say I give, you know, I I teach some people how to do some work. Yeah, I teach the students and experienced people. So let's say there's a work they have to do. A lot of people are used to asking, what next? Oh, yeah, I did this this work, and then after this, what should I do next? Oh, yeah, I made this, designed this page, and then what next? So I think it needs a it needs a better perspective on why are we doing this in the first place? So they ask the question, why? They understand why. And then even before they come and ask what next, they figure out what's next, and then they come and show, here's what I've planned for as the next step. What do you think about it? Is this okay? I I think that's a that's a better way, better perspective. And I see these things missing from a lot of people. And in the AI age, pre-AI age, this was not so important. But in the AI age, I think this is very, very important. People have to be more aware, ask questions, plan their way, what to do next, understand the whole problem. And you know why? It's because all of us now have become a CEO. Every developer is a CEO. Every developer has a bunch of AI agents that have instruct, plan their work, get the work back. Now, they can't have an instructor over them to tell them what's next, next, next, you know, to tell them the tasks. So I think those are some of the things that are changing, you know, the mindset, the way people look at things, and the continuous learning process. You're always learning at a fast pace, always using AI and learning. It's not like you have to take a course to learn, you have to read a book to learn. So those are pre-AI mindsets where you think, oh, yeah, I have to. I'm not saying you don't need books. It's fine to use books, it's fine to take a course. But but the always learning has to be there. We always have to be moving. That's what I said, you know, even before. Instead of using a horse and a cart, now you have a Ferrari car, you just have to keep moving at a faster pace. The Ferrari car is AI. AI is a Ferrari car. Previously, you were using the horse and cart, but now you just have to use the Ferrari car. It's I don't think it's harder. You just have the new tool and you can accomplish much more with the new tool.

SPEAKER_00: 35:31

Yeah. So what if what if somebody hits a a block, a mental block about like what do they what what should they do next? So how do you coach them through that?

SPEAKER_01: 35:43

Well, the simplest way, go and ask AI. You don't know what to do. First thing, go and ask AI. If you're not satisfied with one AI, check with five different models. Of course, I'm not saying trust a model. I I also believe in verifying the information. But if somebody wants to know something, ask five AI models and then then come back. I think that's it. They have a block, just go and ask and then come back.

SPEAKER_00: 36:08

You made that answer too easy. This is gold, by the way. I love this conversation.

SPEAKER_01: 36:15

Yeah. And another thing is using AI for coding is only one part, using it to code. And of course, I think once a developer knows to use AI very well, he can do pretty much anything, even non-development work. Because you have MCP servers, you connect MCP, so you can pretty much do anything. You can connect an MCP that can go and uh just an example, do shopping for you. It can find what people want, can buy and sell. It can do anything you want. Because as long as you have an MCP server, you connect it and your AI can do anything, it can operate your computer for you. You tell it, hey, AI, can you do this on my computer? Just do this. And that's how we have this open claw that is so so popular today. People are using and that that's what it's basically doing. It's an AI agent that you can just send text messages or WhatsApp messages, control it remotely over your phone, and that AI agent goes and does different jobs for you. And that uh basically uses the same thing: the agents, markdown file, the skills, the soul, and they have all these things that describe each thing. You can download these off like apps and install them, and they can do different tasks. So yeah, so there's so much you can do.

SPEAKER_00: 37:27

So here's a here's a thought-provoking question. In a year from now, after all these intelligent people that you're you're working with currently, and those that you could be working with in a year from now, realize that there's a pattern of learning here that they can go and evolve with themselves. What is the role of mentorship in the future when everyone can go and figure these things out for themselves?

SPEAKER_01: 37:57

Yeah. See, even today, everybody everyone can figure these things out, but everyone haven't. And I think that's for example, I've seen I've seen people asking this question, even students. They say, We don't we don't know what to learn. We don't know what to do. I we wish somebody could guide us. We need we need somebody to guide us to tell us what to learn.

SPEAKER_00: 38:22

So that's how the AI that guides them. You're saying use the AI to guide you.

SPEAKER_01: 38:26

Yeah, yeah, that's true. The AI can guide you, but the AI can guide you if you have a goal. But when you don't have a goal, let's say I don't know whether I should be a doctor or an engineer or a scientist or uh I should be into accounting, so they can talk to AI, but AI is not going to tell everybody, hey, you become a doctor, all of you are becoming so if they all go by I want to be, but again, it goes based on a conversation, what you like, and then the AI guides you with it. And I I don't think because there's a way to all these things, and I think everybody haven't figured out the way, and that's where mentorship comes in to guide them, and wise counsel always will have value. So even if all these people do these things, they consult even 10 AI models, but they need some wise counsel. Hey, I've consulted all these things, but what are your thoughts? Because you've been working in this field for a long time. What do you think? What are your thoughts on what I should do? So our thoughts always have value because our experience has value.

SPEAKER_00: 39:33

Yeah, no, that's true. But here's the thing, Godwin. Children these days, they're growing up in a world where they don't know anything other than AI. AI is everywhere for them. And these AI technologies are learning about them. They're interacting with ChatGPT every day, and it's building knowledge. It's building sources of knowledge around like what does this kid that's talking to me do? What are they passionate about? What are they asking questions about? How are they evolving in their thinking as the years roll on? And that knowledge, that context window, right, is is gonna become much, much larger and maybe infinite eventually. Right? So as the as that as that knowledge base of the AI knows us, then it's gonna become the the the wise owl that we can go ask about ourselves, about our deep personal needs, and say, how should I think about this? And it's gonna re it's gonna reference conversations from years, right? So what then, Godwin?

SPEAKER_01: 40:43

That's a great question. I I love that, Jason. So see, I I think see, even though we have AI, AI can guide each one, people can talk to AI and ask AI different things. I think humans at the end of the day need humans. It's one thing, one thing is humans need humans. Okay, that's because we need a family. We need a father, a mother, a brother, a sister, a life partner. That's just part of our life. If we don't exist, these AI agents or have no goal and purpose, they do nothing. It's only we who give them a purpose, and I think we will never lose our value because that's how nature itself works. So let's say we all are using AI. There's a farmer who's using AI and he was using robots to till his field and do everything for him. There's an another guy who's probably having a furniture shop and he's selling furniture. The guy who's selling furniture has to go to the guy who's selling food to buy food to the farmer. The farmer might have to buy furniture for his house and he needs to go to the furniture shop. Now, it's humans dealing with each other, and AIs are just working for us in the middle. It's not replacing uh humans. That's one thing. So humans, human interaction with humans, I think it's always needed. So will an AI satisfy them by being a replacement for a human mentor. Uh human, I think no, because sometimes people want someone to whom they can share things and someone who can really empathize. Because when you talk to an AI, you know that it might tell you that it's empathizing, but you know that it's a program. It's not, I don't believe AI is a program in the sense a normal program. It's a new kind of a brain. That's what I think because the model never existed before. That model is a new kind of file, but still it's it's a computer. It's a computer in the end. It's not, it doesn't have a life. It doesn't, it doesn't, it has never felt pain, it has never felt emotions, it only simulates them. So I I don't I don't think that humans would ever be satisfied with that. That would just be I don't think that that's gonna, even if they do, of course, they're virtual companions and all that. I don't think it's gonna fulfill the goal. AI cannot fulfill the goal where humans are needed to fulfill the goal because that's how the nature is designed, earth and the order of life is designed.

SPEAKER_00: 43:18

Yes, very good. Okay, let's circle back. I wanted to ask about work readiness because you touched on this earlier. If we do this right, what is a work-ready person look like in the agentic AI future?

SPEAKER_01: 43:35

Great question. So I think someone who's work-ready is, you know, I was speaking from a perspective of students. So I'll tell you why students are not work-ready in many cases in the first place, because they're used to studying for exams. They have to learn and they have to write exams, they have to finish their course and get a certain mark, finish their degree, and that's their goal. And so, if they've learned AI, a lot of things they've learned about AI is on the books. How AI works, what is Transformers, how does it do, and they get questions in the exam which they have to write. They're not aware of what they can accomplish using it. They've never been a lot of people, some people are, but a lot of people are not used to setting goals and accomplishing them practically using the using the computer. So though that's one that is one thing, you know, one thing I've been mentoring and training people with, and also their mindset, the way they think, you know, just all that I just said about the decision making, the the asking questions, the way they think, I think they have to think outside the box. And sometimes to think outside the box, they need mentorship and guidance. So, work once they are like this, they think like this. At least in my perspective, for the teams I work, they are a little bit ready to handle it because they know some basics and they can also talk freely, they can be bold. Some people are not bold. If they don't have boldness, they cannot speak. I think they're not work ready because they can't ask anything, they can't communicate. And I think that these things are more important than all the concepts. It's, I mean, they might know Python theoretically, they might know Java, Python, whatever. It doesn't matter. I think the basics.

SPEAKER_00: 45:42

Alright, final question for you, Godwin. What's one piece of advice you give every mentor or mentee in this evolving space?

SPEAKER_01: 45:53

So I was saying when you when you mentor someone you have to do it from your heart. So you do it out of true concern from some for somebody. Because when you're mentoring someone, you're actually helping them. And it shouldn't be out of duty or due to any other intention. Because then that does not that isn't true. You're not really mentoring. Mentorship has value only when you do it from the heart. You you're really concerned about somebody. You really want to take their life forward, and their success is your success. And I I think that's those are the kind of mentors we need in this world. People who really care about the society. I think it takes a heart to do that.

SPEAKER_00: 46:38

Yeah, that I I completely agree. That's wonderful. Thank you for sharing the heartfelt closure. So, in parting, I absolutely want to thank you very much for your time, your wisdom. You've expressed many thoughtful insights, and I'm absolutely positive that our audience is going to enjoy this conversation, Godwin. I know I learned a ton from you, and I want to say that sharing your insights is what the future and the age of AI really requires, and that is the value you're going to continue to bring to society is to help them evolve and learn and grow with these technologies. And I think you're in a great place to do that. So from the bottom of my heart, thank you.

SPEAKER_01: 47:33

Thank you so much, Jason. Thank you for having me.

SPEAKER_00: 47:40

Awesome.