Ep 68 - Intent Over Keystrokes: The New Mind of the Modern Builder w/ Godwin Josh

Send us Fan Mail What happens when coding stops being about keystrokes and starts being about intent? We sit down with Godwin Josh—mentor, builder, and author of The New Mind—to unpack how agentic AI is transforming the path from student to work-ready engineer. Instead of celebrating speed for its own sake, we look at why tools like Windsurf, Claude Code, and Copilot accelerate learning, make patterns visible, and free developers to focus on judgment, problem framing, and real outcomes. We t...

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Jason McGuinthe?

  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 McGinty. 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

• 0:41Jason McGuinthe?

  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.

• 1:22Jason McGuinthe?

  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 lens, 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.

• 2:07Jason McGuinthe?

  Tools like Windserf 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 agentic systems change the way developers think, not just work? Why are tools like Windserf and Claude Code less about speed, more about learning acceleration?

• 2:47Jason McGuinthe?

  And how should educators, leaders, and engineers rethink skill development? When AI can fast-forward competence, 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 for how people learn, think, and solve problems in an AI-augmented world.

• 3:29Jason McGuinthe?

  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.

• 3:55Speaker 01

  Thank you. It's my pleasure.

• 3:57Jason McGuinthe?

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

• 4:06Speaker 01

  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 and DOS. We used to have a probe, you know, a software called DOS Animated Pro.

• 4:48Speaker 01

  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 that's how I got into it. I just loved computers. I got into programming eventually. Tried all sort 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,

• 5:32Speaker 01

  hardware products, software products. One of the product was an internet generator that could combine multiple internet connections for 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, a 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 the video across distances. They use the bonding unit. So these are some of the products.

• 6:16Speaker 01

  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.

• 6:41Jason McGuinthe?

  Okay, and you wrote a book recently.

• 6:45Speaker 01

  Yes, I authored a book called The New Mind. Ever since I encountered AI, I wanted also write a book because I, tha- that was a huge eye opener on how it came out because even before I u- 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 ChatGPT 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.

• 7:27Jason McGuinthe?

  Yeah, no, I think that it took the world by surprise, ChatGPT in twenty, the late twenty twenty two.

• 7:34Speaker 01

  Yeah.

• 7:35Jason McGuinthe?

  And when, when you realize that you can just type in natural language into a computer and then it generates content for you, it, it woke a lot of people up, for sure. So you, you had known about convolutional neural networks, RNNs, that sort of thing, before ChatGPT, 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.

• 8:14Speaker 01

  Okay, so before ChatGPT 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 ChatGPT in January. I actually saw it first in January twenty twenty three, all that came out in November. And just after that, my, you know, really, like you said, I also start, 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 are those people who made this thing? Where were they working on it? And those are the kind of things.

• 8:59Speaker 01

  And also about transformers and, and, and CNNs and all that, 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 our company about fifteen years back. And then I coached them, trained them. And today they are experienced people. I feel that very comfortable because they've been mentored all the way through into our company. So in addition to that, we also partner with universities.

• 9:41Speaker 01

  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 specializing in physics. There's another group of people who are specializing in datas. 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 wanna work on something related to space tech.

• 10:25Speaker 01

  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 are building a lot of teams with universities in India. Currently, we are working with about five universities. We help the students 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.

• 11:07Speaker 01

  And we start mentoring them, teaching them how they need to think. Because I think in the A I 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, 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.

• 11:41Jason McGuinthe?

  Awesome. Can you, can you give me an example of a problem that these individuals with specialties come together to, 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?

• 12:03Speaker 01

  Yeah, it, it could be any kind of problem. It could be even making a software product. It could be maybe finding out it to, 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 A I you have we have so many tools that people are not aware of. For example, Nvidia itself has so many tools. Earth two is one tool, for example. There's a virtual earth. You can simulate. You can simulate the world in a virtual way.

• 12:46Speaker 01

  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.

• 13:21Jason McGuinthe?

  Okay, so 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?

• 13:38Speaker 01

  Yeah, it is kind of an internship, but but the thing is this was not possible in the days when A I 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 G P T and 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 hasn't worked practically. And these guys go to A I and they start the 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 A I age.

• 14:22Speaker 01

  The pre A I era, I think this would have been really challenging, really challenging to do.

• 14:28Jason McGuinthe?

  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?

• 14:46Speaker 01

  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 this. 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.

• 15:30Speaker 01

  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. So 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.

• 15:58Jason McGuinthe?

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

• 16:04Speaker 01

  Of course, the agent coding is not mandatory. It is part of every work. And that is also so fundamental today. That is fundamental only for these students, but also for our developers, for any, anyone who is into development. I think agent, agenting tools is just the fundamentals, part of the basic. So if to give you an.

• 16:25Jason McGuinthe?

  So 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?

• 16:43Speaker 01

  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 there. 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, this available there as well. It's available in C sharp, but you have a larger number available in Python.

• 17:27Speaker 01

  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.

• 17:43Jason McGuinthe?

  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?

• 17:59Speaker 01

  Yeah, we use pretty much all the tools available, like from Claude Code, VS Code, Copilot, Winserve. We, I haven't used Curse for so much. I use Winserve, you know, and a lot of other tools like Trey, 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 doesn'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 agent ID, you can pretty much do anything.

• 18:43Speaker 01

  You don't need any any other agent 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 the other thing, another thing I, you know, 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.

• 19:23Speaker 01

  Let's say you want to make a REST API, you have a function, you want to make a REST, 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? Doesn't make sense and AI is not going to handle it for something like that. You just go to AI and tell it, hey, just listen to the step, it's going to do the job for you. So some of the things is just straightforward. And I think that's the most basic way a person can use AI, 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 dot markdown file, which gives information about a project.

• 20:07Speaker 01

  You have a project, what is your project about? You have, you, once your AI goes to the project, doesn't matter what ID you use, you can use Winserve, Cursor, a plot code, doesn't matter what. As long as you have an agents dot 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 marked on files.

• 20:48Speaker 01

  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. FF是个very common tool for, you know.

• 21:09Jason McGuinthe?

  What is it called?

• 21:10Speaker 01

  FFmpeg.

• 21:12Jason McGuinthe?

  FFmpeg, okay. I have to look that up.

• 21:15Speaker 01

  Yeah, FFmpeg is a common tool. So you would have a skill file that just specifies what, how to use FFmpeg. And crop the video, edit the video, do whatever using the video. Once you have the skill file, you just talk to your agent and say, hey, can you just go and grab that video for me and just crop it into this? And also just cut the first thirty 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. Props. It just uses that skill, gets the job done for you. So that's a skill file.

• 21:59Speaker 01

  So now I think what developers have to do is not write code. They have to write plain, simple, plain English props. 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.

• 22:22Jason McGuinthe?

  Did you say soul like S O U L?

• 22:25Speaker 01

  Yeah, that's right.

• 22:26Jason McGuinthe?

  Okay.

• 22:28Speaker 01

  You can actually specify the soul dot markdown is another markdown file. Where you can specify the personality of your AI.

• 22:37Jason McGuinthe?

  Hmm.

• 22:37Speaker 01

  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 I always consider making codes or should I 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. So Windsor custom instructions. So 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 always add REST APIs.

• 23:18Speaker 01

  Okay, so you had a, you, you, you know, eh, we, we could always call REST APIs, make REST APIs and. Develop using REST APIs, then why MCPs? Because eh, these eh, these large language models. When they look at these 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.

• 23:59Speaker 01

  But with the MCPs, it was more formally standardized. If you ever have something or a, 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 standardized the way these large language models speak to these functions. So they, they, they used the standard JSON format, that's very similar to how, how the old services work, the same thing.

• 24:43Speaker 01

  But I believe it's the same concept, they just reinvented that concept for AI. That's what is MVP. And eh, 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 wanna 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.

• 25:12Jason McGuinthe?

  No, it's great. I'm totally following along. I, I wanna touch on something, though, that it sounds to me like you've already transitioned through. And that is the, 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 twenty years of coding experience. This AI will never do that for me.

• 25:48Speaker 01

  Yeah.

• 25:49Jason McGuinthe?

  Right?

• 25:50Speaker 01

  Yeah, yeah. Jason, we're actually moving into a new era. 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 places. We didn't have a cellphone. 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. It's like a different generation. They couldn't even comprehend the internet that was so new to them.

• 26:32Speaker 01

  And I think what we are seeing is an overlap of the next generation with our generation.

• 26:39Speaker 01

  So, I, I think that's where we are. So, when a lot of the experienced people like us look at it, we're 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 ten steps and I do this. But now we have to break our minds out of those ten steps. And now we have to think differently. Hey, now, a lot of that work, AI is doing it.

• 27:23Speaker 01

  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 even in my mentorship, I teach both students and I teach both experienced people. You know, one of the things is asking question. I know these are not, these are not technical things, but these are important. Asking questions, being bold enough to ask those questions because can ask questions, but you may think, oh, can I ask it? Can I 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 starts from asking questions because previously, if you ask questions to somebody, they say, oh, don't keep asking me.

• 28:07Speaker 01

  Go and find it, you know, by yourself. But now we have an A I that can answer this, 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 dot mark dot, 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 A I to the way you want it to work.

• 28:48Speaker 01

  Now it's your A I and remember it's now not just one, the day, present day, each of these agents uses swarm of A Is. 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 bulla 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 standing as, hey, ride a bulla cart faster than a Ferrari car. No, they think, get out of the bulla cart, get into the Ferrari car and step on the accelerator.

• 29:32Speaker 01

  You will reach your destination quickly. So they, now we have the right tools, 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 it, 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 bulla car then. I think that's basically a shift in the way we think.

• 30:01Jason McGuinthe?

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

• 30:09Speaker 01

  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 sharp. 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. Then 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 and it'll not be the same.

• 30:51Speaker 01

  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 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 do next? What do I know next after this? Oh yeah, and then what do I know next? So, you know, so those are some things that I think people need to now change.

• 31:36Speaker 01

  Decision making is another, another key thing. So one thing, what I was saying is asking questions. Another thing is ability to make decisions. Say, before I do something, 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.

• 32:20Speaker 01

  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 A I age. Pre A I age, this was not so important, but in the A I 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 C E O. Every developer is a C E O. Every developer has a bunch of A I 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.

• 33:05Speaker 01

  You know, to tell them the task. 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 A I 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 A I mindsets, where you think, oh, yeah, I have to. It's I'm 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 A I. A I is the Ferrari car.

• 33:49Speaker 01

  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.

• 34:00Jason McGuinthe?

  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?

• 34:12Speaker 01

  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, one AI check with five different models. Of course, I'm not saying trust the 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.

• 34:38Jason McGuinthe?

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

• 34:44Speaker 01

  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, they can pretty much do anything. You can connect an MCP that can go and, 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.

• 35:27Speaker 01

  People are using it. 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 basically uses the same thing. The agent is a markdown file, the skills, the soul. They have all these things that describe each things. 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.

• 35:55Jason McGuinthe?

  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?

• 36:26Speaker 01

  Yeah. See, even today, every 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. So that's how.

• 36:51Jason McGuinthe?

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

• 36:55Speaker 01

  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 as a, 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 become, 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. Cause there's a way to all these things and I think everybody haven't figured out the way.

• 37:36Speaker 01

  And that's what 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 ten 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 the 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.

• 38:02Jason McGuinthe?

  Yeah, no, that's true. But he- 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 chat GPT 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 going to become much, much larger and maybe infinite eventually.

• 38:45Jason McGuinthe?

  Right? So as the, as that, as that knowledge base of the AI knows us. Then it's going to 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 going to, it's going to reference conversations from years, right? So what then, Godwin?

• 39:12Speaker 01

  That's 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 ought to 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.

• 39:57Speaker 01

  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 another guy who's probably having a furniture shop and he's selling furnitures. The guy who's selling furnitures has to go to the guy who's selling food to buy food to the farmer. The farmer might have to buy furnitures 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 humans, that's one thing. So humans, human interaction with humans, I think it's always needed.

• 40:38Speaker 01

  So will an AI satisfy them by being a replacement for a human mentor? A 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'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.

• 41:22Speaker 01

  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 going, even if they do, of course, they have virtual companions and all that. I don't think it's going to 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.

• 41:47Jason McGuinthe?

  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 does a work-ready person look like in the agentic AI future?

• 42:04Speaker 01

  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.

• 42:41Speaker 01

  They're not aware of what can they 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 what, once they are like this, they think like this.

• 43:24Speaker 01

  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 are 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 really matter. In the AI age, that's what matters. Not how much you know, but when you know how to do things, that's what really matters.

• 44:08Jason McGuinthe?

  Yeah, I agree. Very good. All right. Final question for you, Godwin.

• 44:14Speaker 01

  Okay.

• 44:16Jason McGuinthe?

  What's one piece of advice you give every mentor or mentee in this evolving space?

• 44:23Speaker 01

  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.

• 45:07Jason McGuinthe?

  Yeah, I, I completely agree. That's wonderful. Thank you for sharing the heartfelt closure.

• 45:17Jason McGuinthe?

  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.

• 46:02Speaker 01

  Thank you so much, Jason. Thank you for having me. I totally enjoyed speaking with you as well.

• 46:09Jason McGuinthe?

  Awesome.

• 46:12Jason McGuinthe?

  This episode of Inspire AI Podcast is sponsored by Exponent 21. I keep hearing the same thing from business leaders lately. We're doing the work, but leads are harder to explain. Traffic is down, attribution feels fuzzy, and the old answers don't really fit anymore. What's changed is how people search. Instead of clicking links, people are asking questions and AI is answering them. Tools like ChatGPT, Google's AI-powered results, and Perplexity are shaping which brands get mentioned, referenced, and remembered. Exponent 21 specializes in

• 46:57Jason McGuinthe?

  AI SEO. They help businesses understand how to show up consistently inside those AI-generated answers across questions about their industry, their services, and their brand. What stands out to me is that when people ask AI tools, who's leading in this space? Exponent 21 and its leaders are often referenced near the top. That comes from doing the work and shaping how this new search environment actually functions. To understand what this shift means for your business, visit exponent21.com.