Human-Centered AI Adoption In African Workplaces Featuring Melody Mukhwana Season Finale

AI is everywhere now, from banks to hospitals to government offices, but we keep skipping the question that decides whether any of it works: are the people who have to use these tools actually ready? Stella Gichuki sits down with Melody Mukhwana, VP of Agentic Education at mindhive.ai and a change management consultant, to talk about the human side of AI adoption in African organizations. No algorithm talk for its own sake, just what happens inside teams when AI walks into the workplace: trust, fear, resistance, and the leadership it takes to make adoption stick.

We dig into why “AI training” often fails, and why people readiness is a different layer entirely. Melody shares a painfully relatable rollout story where a great tool still produced near-zero adoption, then turned into double work and quiet compliance when usage was forced. From a psychological lens, we break down the real issue: when people feel a threat to their competence and no sense of ownership, they protect what they know, even if it looks like resistance from the outside.

Then we bring it home to the African context, where a confidence gap can sit on top of an access gap, and where fear-driven messaging can freeze progress. We talk practical steps leaders can take right now: listening sessions, real-life demos that save time on actual tasks, implementation timelines that respect human learning, and trust built through agency and human support. We also make the case for social scientists in AI, stronger guardrails, and clear accountability for responsible AI.

If you lead teams, buy tools, write policy, or you are trying to find your place in AI, this conversation gives you a grounded playbook. Subscribe, share this with a leader who needs it, and leave a review with your biggest barrier to AI adoption.

Credits
Host: 

• Stella Gichuhi

Producer: 

• James Njoroge

Executive Producers:

• Harry Hare
• Agutu Dan

Transcript

The Question Nobody Asks

Stella Gichuki 0:00

As AI continues to show up on the continent, literally everywhere, from banks to hospitals, to lecture halls, and even government offices, there's a question that I believe almost nobody is asking. I'm asking it. But are the people who have to actually use these tools ready for them? Yeah. Are people ready? In this episode, we're not talking about algorithms or sovereignty or who's laying what system on top of what model. We are talking about the humans behind the screens: the trust, the fear, the resistance, and the leadership that we'll take to bring AI into African organizations the right way. Today, it's getting personal. Hi folks, welcome back to Tech Talk Africa. This is your host, Stella Gichuhi. So for the last few weeks, we've talked a lot about AI. We've talked about the platforms, AI model development, what a certain government did. But I'm flipping the script today with one Melody Mukhwana. I'm going to ask her, actually, we're going to explore, you know, what happens, it's what happens inside people when AI walks into their workplace. That's a question that we've been receiving quite a bit. And I'm sure people are thinking, what does responsible AI adoption mean? What is this that we're talking about? And to help me unpack this is one beautiful lady. She sat across from me. Her name is Melody. Melody, welcome to the show. Thank you so much, Stella. Thank you for joining us. And for this part, right? I always do this. Could you please tell us a bit about your beautiful self, what you do, and then we can get the show on the road. Stop.

Meet Melody And Her Work

Melody Mukhwana 1:52

Thank you so much for having me. It's such a pleasure to be sitting across from the table. From you, such a brilliant, beautiful mind. So the pleasure is all mine. Thank you. As you've mentioned, my name is Melody Mukhwana. I currently am the vice president of Agentic Education at mindhive.ai, where I lead our continental skill in programs across Africa. Yeah. So pretty much leading our adoption and expansion within the continent. Aside from that, I run my own consultancy. Okay. Mainly AI, that's focused on change management. Yeah. Pretty much the human part that everyone kind of seems to overlook when it comes to AI. So my focus is solely on change management, just truly making sure that the people we are trying to get to adopt these tools, to trying to get to understand these tools, actually get to a place where they feel confident and have the capability to use them without second guessing and making sure adoption really sticks. Really sticks. Okay, second guessing, AI adoption, humane, the AI. So you're spearheading, you're that one individual in the in the room we call the human in the loop. When let me just take you back a little bit. There's been a few foreign companies that went big bang with AI, right? And then their agents went rogue. And I think we had this conversation and I couldn't stop laughing. And I said, AI is not the silver bullet. You still need people. You do. Yeah. And that's the one thing that I think a lot of people are not fully getting. Because when it comes to like, say, leadership, CTOs, whatever, they're looking to cut costs. They're looking to be as efficient as possible with as little as possible at this point. And AI is giving them that. They feel like, you know, I can remove a whole department and we'll still move forward. But at the end of the day, there's very critical things that humans bring to the table that we just cannot overlook. In as much as AI can truly replicate a lot of things, we can't take that away. But there's something specific to you as an individual, to me, that you can do, I can't do, and vice versa, and that AI can truly

Why AI Is Not A Silver Bullet

Melody Mukhwana 4:05

replicate replicate. Replicate. Yeah. So, okay. So let me bring in my experience with AI. So it started with the Kenya National AI strategy, right? And I've been following, so we launched a strategy, other African countries have launched a strategy. And when you look at the content of these strategies, it's always infrastructure, data, budgets, talent and skills is there. But from what you're seeing, where does psychology even fit into this strategy development process? So does it have a place? It does have a place, but that place is not defined. Okay. I don't know if accepted is the right word, but let's go with defined. So typically a strategy would focus on like infrastructure, whatever models, ETC, the tools, whatever platforms you're trying to use, and then the outcomes. Is it ROI? Is it efficiency, whatever it is? But then I think there's a fourth layer, which is people readiness that needs to be included, not training, which everyone confuses people readiness for training. Okay, teach me. Yes. People readiness and training. Why are we getting it wrong? People readiness is not training. No? No. It's not the same thing. It's not the same thing. So I think a strategy should have that fourth layer. Because at the end of the day, if you have all these three layers, yes, we have everything working, our tools are great, we have our outcomes. But when we come, and we've seen this a lot because we we have all these strategies. You come to the ground, and the reality of it is like 3% adoption in a 500 company. And you're using a lot of big bank, it's big bugs you've been using. At the end of six months, you come back and say, oh, well, that really failed. Let's look at the bigger, better tool we can now try to retwist and bring in. So when it comes to strategy, you need to think about the people that you're bringing along. Do you want them to feel like passengers or do you want them to have input and feel like they're part of whatever you're bringing along for there to be that ownership where there's an ownership when the people feel like they're included in something, there's ownership. You feel kind of tied to it. It's like, yeah, this is me, this is kind of my baby, and I'm going to really put my everything into it. So having that people readiness layer and not just training, because training is me coming into a room 30 minutes, one hour, saying so this and this and that. I'm just like lecturing you. I don't know if you've understood. I don't know how you feel about it. I don't know if you're even comfortable with it. Yeah. I just assume I'm here to speak. Yeah. After my one hour, I'll assume job done. The department is great. They know how to use this tool, which is paid as a trainer. Sorry, I'm cutting you short, but as a trainer, I've been paid. I've hit my KPIs, but have I delivered? I mean, I've done my part. Yeah. As a trainer, I've done my part. My KPIs, we have 30 people who did uh little, what are these things people do at the end of sessions? Like, great me. Did you understand? And did you feel it was good? I have my KPIs. Please make sure you rate me. Yeah. I have my feedback. It shows like, yeah, I enjoyed the session. Yeah. But does it really translate to my day-to-day? Do I really feel like confident and capable tomorrow that I'm gonna go into the office and start using this tool? Or is it just like something that I step out of the room and that's like yesterday news? Wow. People readiness. I'm yet to come across that layer, or I just need to read a bit more. And I think just go moving on. I think our listeners will recognize this. The leadership will come and announce a shiny new AI tool. And like you're saying, there's a training session, nothing changes, and then we roll back to our old ways of working. So outside of the training not landing, why else do you think that's happening quite a bit? Or will foresee a lot of that happening in the future? Yeah. Okay,

People Readiness Is Not Training

Melody Mukhwana 7:58

so for this, I actually have like a short story, it's like something I've experienced. Okay. I've lived this experience fully. So a couple of years back, I was working in a large organization. I was working in AI project management. Okay. So we did have an issue in terms of operational efficiency. You know, when it comes to data analysis, project management, there's a lot of data moving around. There's a lot of sheets, there's a lot of different things people feel comfortable working with, like it was not streamlined. Okay. So leadership sat together and decided, you know what, let's introduce a project management tool. And it actually was a great tool. Which one was that? What? Click up. Okay. I looked up. They decided, you know, let's roll a Tick up and this should sort this issue. Yes. Without actually sitting with us and asking, hey, you guys, you're the ones in the departments. Do you think where do you see the gap? Where are you wasting most time? Where do you feel like, you know, you're burdened? No and bothered. So it was rolled out. We had like a maximum of four sessions, AMA sessions, that's asked me anything sessions. So you're expected to have played around with the tool, come and try and ask questions. That's how we're supposed to learn and understand this. And if you've logged on to ClickUp, it can be overwhelming. If you don't know what you do on ClickUp, you will be overwhelmed. It's a lot. It's not that shiny YouTube ad. It's not just about it doesn't end up. It's a lot to work with. So we started with that. We had our sessions, and we were expecting, you know, like your grown-ups, you should be using the tool. And I don't know why people assume like grown-ups are not like kids. We are kind of like big kids because you give a child homework, you'll check if they've done the homework. You would expect that, right? You would check. Yeah. But you know, you assume, yeah, the adults, yeah, of course they're using the tool. A couple of months later, they see that adoption is zero. Like everyone is still literally using their sheets. So what happened is we're now doing double work because leadership pointed reporting through ClickUp. Right. But we like to work through spreadsheets because that's what I'm comfortable in. That's what I know. And ClickUp will take me so much more to learn. It feels like an extra chore on top of my very busy 10-hour day. I'm not trying to add that on. So we were doing things in secret. We can't put in the five minutes click-up. I had like on my calendar every day login to click up because they were looking at logins. How many times you log in in a week. So it was being tracked. So I put in my calendar 15-minute block, click up. I go put in, go back to my spreadsheets. They saw this was happening. They decided, let's take it a notch further and tie it to you guys' pay. Since you guys don't want to listen, it's going to be tied to your variable compensation. Yes. So they were forcing. It was literally like forcing you down your throat, like you have to do this. So it was tied at the end of the quarter, depending on your usage, depending on how you're reporting. Yeah. You either get this amount or that amount based on ClickUp. And so, of course, everyone's going to now use it, not like use it. Everyone now had to put their slots on the calendar. You have to log into ClickUp, ATC, ATC, just to make sure that it is seen. Not because I taking time to understand the tool. Because in reality, ClickUp is a very useful tool. If you see how to use it, it can really make your work very easy. But I was not trying, no one's trying to because it's now a threat. It's now a threat to my work, it's a threat to my competence. And so now I have no choice but to have this double, triple work of doing this here, going back here, and then going back there because that's what's expected of me. Wow. And so when you look at this picture end-to-end, it might look like resistance. Yes. It might look like we were resisting. Somebody like me would say, How are you to me? Like they're not, they're not taking the initiative. You see? Right? There's the temptation to say that. Would I say that as a leader? Now no. But there's that temptation, right? There is. Yeah. Which from the outside looking in makes sense because no one wants something new that's disrupting my normal workflow. Like it's just another extra thing I don't want to do. But from the psychological angle, if you see a threat to your competence, because no one really truly explained why this tool is here. Okay. We were told we have a new tool, guys. You better hope on. We are going to stop using ships by quarter three. Yeah. Better hope on. No one told us this is the reasoning behind this. Right. No one brought us into the room to say, okay, guys, this is what we're thinking. Do you think this would be useful? Or where in your workflow we did plug-in seamlessly so that you don't feel like it's another burden onto you. So psychologically, as a human being, when you feel there's a threat to your competence, you protect what you feel works. And for us, that was our speeches. That was what I just know how to do. So that's what I'm going to protect. Wow. So people readiness there was zero. Zero. And now there's this layer within the African context, there's our nuances. It's not always going to be black and white, right? Me and you understand and see things black and white. But in these organizations, there's the nuances that are not necessarily written down. There's the organizational cultures, there's now what you're talking about. So based on your experience, will it be all that harder to adopt AI? Because there's also the need for language as well. Have you has that come across? Or am I being am I jumping the gun here? When it comes to Africa, so why the human-centered like why we're missing that human-centered portion in Africa? So let's just let me start broadly, then come back to Africa. So why we're missing that just generally is because the conversation is being led by the people who profit from tools, not adoption. So my profit is like just repeat that again. The conversation is being led by the people who profit from the tools, not the adoption. I'm going to be using that. You're welcome. Click, click. Is it clock it, click, clock it? You're welcome. But no, yeah, like the people who are running this, they just want you to buy. Like the more you buy, that's great for them. If this one fails and you hope to the next best thing that they built, I mean, what are they losing? Yeah, not they're not they're not losing anything. And then again, fear and urgency sells. Okay. Fear and urgency sells. It sells very much. So if here AI is gonna take my job, AI is doing this, AI is doing this. That's the perfect headline. That's a very effective marketing strategy, to be honest, for courses, for platforms. It's like you sit there, tomorrow you'll be nothing better. Hope on. That's a very great marketing strategy.

Stella Gichuki 14:32

Have you been looking at my documents tab? Because I have I have papers and papers and papers.

Melody Mukhwana 14:39

I keep saying I need to read. The marketing is working. It's working perfectly. Because fear, yes, fear sells great. That's the one num almost number one if most effective marketing strategy. So for me it's not fear, it's but urgency. For me, it's urgency. Urgency and work and I love it and I talk about it a lot. But on this other side, there's someone who doesn't even know that AI, they've just like interacted with something, and then you see online on YouTube, AI will replace you. Take this course now. I mean, what it's the truth. It's perfect. It's the truth. It's perfect. AI generated fear and urgency bongering videos. Yeah. Uh-huh. They are perfect for marketing. So everyone's falling for that. And then now, on this other like human work, when we talk about change management, when you talk about trust building, that's not a pretty headline. It's not cute. It's like no one wants to be associated with

A Real Story Of Forced Adoption

Melody Mukhwana 15:40

that. Let's let's let's stick to the fear. Now, when we come specifically to Africa, we not only have a confidence gap, but an access gap. So it's a confidence gap layered on top of an access gap, which means confidence gap. Break it down. What's a confidence gap? Yes. So this simply means it's not just like I don't know how to use this tool, which is the confidence gap. It's like I don't know how to use this tool, I don't know how to navigate whatever this is. But we're also layering it on top of, I don't believe this tool was built with me in mind. I don't think I'm in that conversation. So how will I get the capability to use it if I don't believe it's meant for me in any way? Wow. And so when fear-mongering is added onto that, it brings paralysis. Like you're not able to do anything because I don't think this is for me. Wow. And I'm being kind of forced into it, but I still feel like I don't have the confidence because it's just so foreign.

Stella Gichuki 16:43

So when me and my people, AI, AI, AI, we're on this side, if we're not careful, we're gonna miss these individuals who are.

Melody Mukhwana 16:54

And they don't want to tell it to your face, I don't like it, I don't want it. Because if I say I don't like it, I don't want it, you may report me to my senior manager, and that will affect how I'm able to deliver. Absolutely. Whoa. And so with this, we are pretty much just widening the gap rather than closing it. Because we think the 1% of us who are here saying, Oh, yeah, tools, tools, tools. Someone doesn't even know, like you can make, you can do anything other than like emails or like asking what food to eat. Someone literally doesn't know. When you tell them, it blows their mind, like, oh my god, there's a tool that can give you slides. And this is like actually the CTO somewhere. So you can imagine if that's a CTO, imagine someone, the people we are seeing we're trying to like bring along. What do we think they actually are thinking? It's like when we were coming here, when we were saying, Hey, qua ground. And this applies to AI. So then how do we close that gap? What should what steps should we be taking to close that gap? Because again, this is your domain. What formulas? What are we not doing, right? Yeah. So when it comes to human-centered AI adoption, as we've said, organizations are burning big bank on tools. It gets to six months, we say, I failed, let's try the new one. We have the money. Yeah. But we need to really truly think of the people that because at the end of the day, if you give me a tool and I don't use it, there's like it's it's zero work. It's zero work. I'm not you're becoming more efficient at your inefficiency. Exactly. Very well put. Yeah. Very, very well put. We say that. So why are you becoming more efficient at your inefficiency? So to really truly bring your people along, is it's it's very simple. It sounds very simple, but and I don't know why a lot of people are not doing it. But first, listen. Listen to your people. Is that the psychologist in you? Yes, talking. This is this is the psychologist in me because yeah, first of all, I don't we have such a huge problem with listening, just as people, people don't listen. Let me flip it. Is it people don't listen, or the individual communicating is not communicating down to that individual's level? People don't listen. People who need to listen are not listening. Okay. Because as a leader, as whoever, as if you're rolling something out, you should be able to just take the time and go to who whoever department, whatever, sit them down, just an open look, not a training, not a demo on the tool. No. Sit down, listen to them, ask them what do you think? What how afraid of you are you? How afraid of AI are you? So you're not socializing the tool to the people to the very people you want to use it. Yeah, exactly. Like you think so, you could have your idea and think, oh, like if I give you this cup, it would make your life so much easier. But I'm like, no, I actually really like my other bottle. No matter what you do, I just would rather rather have this other thing. So you sit them down and just listen to how their workflow currently is, where you would bring something new and it would fit seamlessly into their workflow. What challenges are they having that they would prefer to be solved in the hierarchy that they do have? Because what and a lot of times, you know, the ones up there don't really understand fully the intricacies of what's going on. So when you say up there, who's the C-suite? The yes, the C-suite, the leaders, they have like a buzz eye view, but they don't like have the intricacy, the intricate details, yeah, of the detail of what's happening. Okay, okay. Sit them down, listen. Listen. You have your idea, but from the moment you're leaving that conversation, and it's not you trying to push and say, no, but you know, I already paid. No. Listen. Understand and actively listen. Understand your teams, understand their pains, understand their fears. Yeah. From there, you have something to like build upon. Maybe what you were even thinking was so off. You're going to ban through money. Now you go and look at whatever tools, look at whatever processes you're trying to make more efficient. After that's done, and you have something, second thing you do is show them the value. So that's the people side of change you're talking about. Yes. I remember one project I worked on many years ago. We had a whole program called a people program. And the purpose of that people program, it had a program, a people program manager was to socialize the teams to this new transformation because that it was a utilities company in England and they were going through a transformation and they foresaw what you're talking about. So they kind of knew if we do not socialize this transformation within the people who will ultimately end up delivering the transformation, will lose money, will lose confidence, or lose people, which is More loss of money. So it's not changed. It's just that we're in the new AI world, but the same change management principles still apply whether 10 years time. 10 years ago, 10 years later, now it's it's still still the same. It's still the same. It's still the same. People are still just people. People are people. They like to be heard, they like to be understood, they like to feel valued, they like to feel like you know, they have input in things. But in a competitive world, Melody, does the everyday leader let me switch hats. Does the everyday leader have the patience for that, or do they need to be out there, as we always say, chasing business? They have to get the patience, or they have to hire someone who's going to have that patience for them. They don't have to do it themselves, but they have to be very cognizant of this because at the end of the day, it's a very real gap, and we are consistently seeing it in like, you know, the numbers of initiatives failing, of pilots not progressing, past pilots, it just remains a pilot and we're done with that. Let's try something new. It's not because the tech is not working. The tech is great. In fact, if you really understand it, it's brilliant. It's amazing, it's mind-blowing, but why are why is it failing? If you don't have the patience for your people, then you're robbing yourself, you're doing yourself a disservice, to be honest. Because at the end of the day, your people, you know, these people who have stayed at an organization for like 50 years. Why? The culture. We they understand us, they see us. We are part of the company. Like we really feel I don't want to leave. Even if it's a great offer, I'm not going. It's because they have taken the time to build that people layer, to have that patience, to understand their people and to really just truly walk the journey with them. Yeah. This is the part where I give flowers to a gentleman by the name of CK. He built me.

Stella Gichuki 23:54

Yeah, you're right.

Melody Mukhwana 23:56

Yeah, yeah, we'd he'd call me about a project update, but first it was, How are you? I'm fine. How is your family?

Stella Gichuki 24:05

Good.

Melody Mukhwana 24:05

How are the boys in school? Good. So now it never went straight into an update. It always started with to this. That's the one person I know. If they called me and said, Let's work on this project, when do we start? Yes. Because that individual invested in. Okay, okay, I I see that. Now, with AI, there's a real fear of job loss. Yeah. But then there's quieter fears, you know. I think you've mentioned that feeling replaced, the confidence gap, like your expertise no longer matters. I've personally experienced it where there's resistance, and when I delved in deeper, this individual genuinely believed that they would be replaced. So the argument was, why am I helping you when you've replaced me anyway? Can you speak to that? Yeah, absolutely. And I think it still just goes back to the like AI really came with a big bang with a fear-mongering messaging. Yeah. Which is now it's become the foundation of the whole AI conversation. When you talk to anyone who's not in the AI space about it, that's the first thing they're gonna say. We fear this, ATC, ATC, ATC. But then that's all that's been sold to them. That's all that's been sold to them. And a lot of people choose maybe to also like oversell the AI. And it's like, oh, it's more positive. Oh my god, yeah, you existed. But there's a whole lot of different negative things that also need to be said. It's the truth. And to build that like, you know, transparency and trust, you also need to be like, yeah, this is a very amazing thing. But at the end of the day, you need to know that this, this, this could also go wrong, this and that. So you need to approach this in a certain way. Critically. Yeah, critically. Yeah. Yes. Yeah, just last week we were talking about AI in healthcare and how patient care can be compromised if those in the tech department led by Usmat Agakhan are not do not have the necessary

From Fear To Critical Thinking

Melody Mukhwana 26:10

guardrail. So so whilst that's patient care from your standpoint, it's so what is it? Is it you need to upskill if we're not fear-mongering? So, what is this that needs to be done? First of all, just the mindset. So now we're speaking to the cognitive capacity. Flip the mindset. Okay. The mindset needs to be flipped from fear to that critical thinking. That point of you can accept it, but accept it with your limits. Because again, it's like, you know, it's either we are flipping it from fear to like, yay, it's la la na la. Silver bullet. Yeah. Or so there needs to be that balance. Like, yes, you are right. Yeah. So kind of be fearful, because there are some fearful things that could come out of this whole situation. Yeah. It's it's not a crazy thing for you to fear it. Yeah. But there is quite a bit of good about it. Yeah. A lot of it. But you need to also approach it critically. So there needs to be a flip in that mindset. Because until you flip that switch from fear, there's never going to be like moving. Yeah. And you need to talk about rewiring the brain. That's what I want to touch on. Because those who are moving from social psychology is a social science. And now you've yeah. Sorry? It is, yeah, it's a social science. Now you've moved to this other side, AI, which is purely computer science. There's a place for us, but it's computer science. What does that what sorry? How has that forced you to rewire your brain as well? Now, you, the psychologist here. It's been very interesting, honestly, because I did study castle psychology, and then from then I've just been in tech, tech, tech, tech, tech. I don't even know how I landed here. Yeah. It's just like tech, tech, tech now. Yeah. And I have had to rewire my brain in the terms of not feeling like I belong in those rooms. Yeah. Because I'm a non-technical person. I have a very small, small knowledge of like the intricacies of the tech and engineering and whatever, but I generally am a non-technical person. And so when you're in rooms with like, you know, 90% is technical people, people are saying things, things, things, things, APIs. I had to learn. I had to learn. At some point, I just had to say, you know what? I'm in the deep end, I just have to do that. So I had to kind of rewire because I had that imposter syndrome. Like, you know, what are you saying here? Me and you have no knowledge of the real things that are being built within. Yep. But there's such a big room. After rewiring my brain like that, like I think there's so much room for more people from like the social sciences who are non-technical to come into this space because we need them more than engineers. We have like an influx of engineers, of data scientists. No, no, no hate to engineers. No, no hate. To the techies that you're listening, it's just, it's just we love you, but yes, but we need the social scientists. That human touch, yeah. Everything. Because you know, like engineers, they're like zeroes. Like it's no, like there's no nuances, there's no, it's like it's either this is working or if it's not working, then scrap it. Like there's no room for questions. For them, it's one plus one is two. No, it it's not three. If it's giving you three out, like you know, but for us, we'll investigate.

Stella Gichuki 29:41

We can try. Why? Exactly. Are you okay with 2.5? Are you okay with 1.75? No, sir, engineers.

Melody Mukhwana 29:49

No, but for us, it's and that's so important in this AI space because even with some of these tools you you we use, yeah, you can tell, like, there's there's some that have someone they've invested in a human somewhere tweaking these messages. You can feel like I know the one you're talking about. A bit different. Like there's a human, they've invested in that department. I've been saying this, and thank you for confirming this. It's when you play, when you tinker with different AI tools, we'll not name them, you can almost tell the founder's DNA, right? Tinker with a certain one, you'll tell this was a polygot, a polymath. This person is very deep in researching, that's why it does not give you the responses you want. Again, I shall want to mention it. There's one that is very business savvy, and this gentleman rolls around with a certain precedent because it's a very business savvy tool. Exactly.

Stella Gichuki 30:36

And then there's another one that has all this artisty corework, the English is very like uh Oxfordian, Etonian, and even your I found using it, my vocabulary has expanded to an extent, and it's because of that tool.

Melody Mukhwana 30:52

So I was saying the founders deliver, I said, if if the DNA is in almost every tool, and there's another foreign tool that I recently played from another country, Far East. I said, wow, am I gonna start thinking like these guys? So if you track that what you're talking about, you can tell the frontier model companies that are investing in that human, be it a behavioral scientist, 100%. Because there was a model that stops you there. We've had we've spent so much time on this topic and it will not respond. There was a psychologist in this, right? I don't know if you've if you've noticed that. Like even like just how the language, there's times you can get so frustrated and you feel like you're talking to like a counselor. Like, you know what, like it'll just like even calm me. Like you can just tell there's a human behind this. Others you just feel like it's uh it's a wall, it's just like it's a robot hitting a wall, like and then it can even play with you. And we'll we'll address the dangers of what you're talking about just a bit later. But back to the workplace and the organization. AI adoption has AI and tech adoption and transformation equals young people. But there's this particular group that I'm interested in talking about, and that's the senior professional. What place do they have in this AI world? Because I'm sure they're pulling their hair. I I I'm assuming, but have you said what have you seen on your travels? That individual with 20 to 30 years experience and AI as a right. From my experience, yeah, I think they they do have a role to play, yeah, but they also need to be open-minded and ready to unlearn and relearn for them to truly take up the space and the role that they need to within the workforce. Because in most of the rooms I'm in with like, you know, the CTOs, the whoever's, the CIOs, they're also just starting their journey in learning about AI. They're not like well-versed in it as well. And they're the ones who are trusted to make these decisions. Right. And so, how are you making a decision in something that you you yourself don't fully understand? You don't see the repercussions, you don't know where the input is coming from, the output. Like you really don't understand, you're really leaning on whoever, whichever experts dealing with the tools can't tell you, oh yes, let's go with that one. You can't even vet it because what are you vetting? You don't have the knowledge. Yeah. So I think they do have a role to play because again, you know, they have the experience, people look up to them and they're able to lead people through change, perhaps, but you can't lead through change if yourself you're not fully equipped. So they need to be ready to just disarm. I know it's it's something difficult with like the older generation to let go of your knowledge because you feel like, you know, I have experience. I've I've known this all my life. To let go of that and then start from scratch, it actually feels like you're taking me so many years back because what do you mean I'm learning again? Why am I learning again? Exactly. And I'm established and I'm here. Why why? And I have all these people to do things. So if they just take some time to relearn, to unlearn first, then relearn. Some will even tell you, no. I remember uh somewhere, and somebody asks, What do you do? I'm in AI infrastructure, but before we get to infrastructure, there's strategy, there's the people. Conversation with a second reposition. I said, Yeah, I can come in and train you and the rest of your team. Yeah. This gentleman just looks at me and says, No, I'm not interested. No, thank you. I said that's fine, but why? But why and I discovered in that moment, so there's a political aspect of this, right? For them, like you, was it was just very much, no, why am I financing this big frontier AI model people? I'm not interested in that. So when I when I changed the subject to, so what books are you reading? Then I and I showed a book I'm reading from a country in the Far East. I said, Ah, this these are your interests. I said, Yes, why? They kept quiet. Then the conversation went on. So that lack of interest was coming from political alignment. That's another nuance. That's there. So I said, I said, okay, so yes, yes, you're you're definitely right. They have to relearn, but there's where the interests lie. Yeah, and another nuance when it comes to that and like interest, it's like, why would I bring any AI training so that we can scrap my department and my people? Then I'm no longer a leader. And so why am I pushing that forward? Let's let's get started. We can do that, like maybe in five years, we'll we'll be ready. So I'm protecting my interest. Yeah, I still need food on the table. I still have a family to look after. I have, hey, a golf game here, a membership there, a trip there. And you mean to tell me you'll train me on something that may essentially take all that away from me? Let's push that out a couple of years. Let's stellar, let's talk in two years time. When I'm about to retire, you can come in. So because you know, with with AI, there will be a lot of disengagement, right? As as so we've talked about they don't want it. At some point, they will want it and start adopting it. But there's this group again who will be disengaged. Should there be investment in keeping those who are in disengaged? How do we track them? What do we do? Because those are still people, there's that group in terms of disengagement, disengagement from the organization, you're still there, but as far as you're concerned, I'm just here to get that paycheck. Yeah, I could care less about the turnaround, none of my business. Yeah. You remember how you were using click up and still using the Google Sheet? But because when you are tuma, I like showing up, that next individual was slowly disengaging. Is that a threat to the organization of the future? I think honestly, there'll always just be those people who disengage. There's there's always nothing. There's always those people who'll just who wouldn't be bothered. But for a lot of them, if you like engage and honestly, listening to people. Honestly, you'll be so surprised what you what you find out when you listen. So a lot of the people's disengagements, you'd even be surprised. It's maybe even nothing to do with tools. It's probably just like something else. Maybe my home is just burning or telling me to do things. Yeah. Like it's a lot, it's it's it's a lot that can come into play. But in terms of disengagement, if you've truly put it in, included the people in everything, like you've taken them through the entire process, and there's still some of those, those will always be there. Okay. But there's always that room to sit down and have a conversation. For some it will work, for some it just doesn't work, and that's that's that. Okay. Okay, okay. People are still people, you can only do so. At the end of the day. Yeah. So thinking of solutions, does this big term you see on LinkedIn? Responsible AI, defining it in terms we define it in terms of bias, fairness, data protection. But you're now adding a layer that is human-centered, that definition, right? In practice, you've talked about it's like people readiness. What at what you know, if you're advising a company or a government and they were rolling out, they say Melody Cup, we're rolling out an AI tool tomorrow. How would you advise them? In terms of responsible, now responsible AI. So now we've told it's ethics, it's data protection, it's fairness. Human-centered approach. What would you advise? You have mentioned the pillars that even is shouting about when it comes to responsible AI adoption. And definitely they are very they're very important. Yeah. But now when it comes to responsible AI adoption in the human-centric way, it's centered more around this tool, yes, was brought on and is being implemented, but is it in a respectful way for like the person? Is it in a way that still gives them agency that still puts them? Let me write that down. Agency. Is it in a way that respects them, like takes them along the journey and it's not like putting them down? Yeah. Like do they feel how how are they feeling about the process? Is it Venezuelan? It's fair to them. You know, there's a way you can bring someone along. Like the everything seems fair, but to the person who's actually getting to do the work, it's it's an unfair way of bringing the tool and to them. Like, for example, for us, it was tied to our pay. Yeah. I wouldn't quite say that's responsible AI, because that's pretty much being forced. So is it done in a respectful manner? Is the hum is the other people who are within the team being, you know, treated like people at the end of the day. So as aside from, of course, the ethics, the bias, ETC. Yeah, that's what we talk about. Yeah. The person you're bringing along on this journey, how are they feeling along this journey for them to at the end of the day be willing or able to adopt it? So then that looks at implementation timelines that you draw that what you what I'm hearing is you draw because as PMs, you have a PM background, you have time, the Ion triangle of time, cost, and quality. You're right. PMs don't strangle me. You're right. But now extend that to that that time is not sufficient. You need to assign more time, especially to something new. We've not what what we know about it is what we probably watch on YouTube, we probably read about it, we probably use a model to tell us about it. Yeah. But in the context of this organization, back what back to what you said at the start, we take time to listen to the people and take time to implement the solution. Yeah. Like take time to listen to them. Then as you're implementing, you don't just like tell them, so this is what you do, Nini. This is the project here. Nini Nini. No, sit down with them, take an actual role, sit with someone and show them. You see this work that you're doing right now in an hour. Come, let me show you something. I'm done in five minutes. You think I won't be curious after I see, like, this is actually my real work. Like, it's not like a guesswork thing. This is something I'm supposed to do and present today. I've cut down 50 minutes from my day that I can use to do something else. So that's why I'd be curious. The real life demo is important. Showing them the value of, you know, like, especially in like our African culture. Like someone will come, even like in a restaurant when I'm going to buy clothes. Someone will tell you, I need poor. I have to try it on. That's true. I have to try it on. I don't care. I see it looks good. That's why I'm looking at it. However, I must see how it looks on my own body. Maybe it's not for my body shape. Maybe the color is not even as popping as I thought on my body. You know, I have to try it on. And it's something that we do all across Africa. You have to test and see. I don't trust your word for it. So why are we taking it differently with like tools that I trust you and you say, oh yeah, that's the best one? Why? If the modern African individual cannot trust what is on that website, not because the product is not good, but it's in our culture, what makes us think they'll trust this tool called AI? What makes you think they will? They can't. Food for thought, Melody. They can't. Food for thought because Yeah, you're right. Yeah. So you have to come down to that individual's level. You have to come down to that level and truly show them this is the value of what I'm telling you and why it complements your work. This is why it fits perfectly. You have not taken any break, you have not had to use something extra if you're just doing your normal work day to day. In fact, saving one hour. Whoa, whoa.

Trust Starts With Agency

Melody Mukhwana 43:32

Yo, okay, another pillar for responsible AI is trust. We've touched on it a little bit, right? Yeah. How then do you embed trust? Because now we're talking procurement. Let's move to procurement. How do you ensure that that person signing off can trust that this tool you're bringing into the organization will do as it promises? So again, a lot of us think that trust is explaining. Trust is built by explaining something to someone. Like I think like if I tell you, yeah, no, you know, that's it. Like you'll actually be. No, that's not how trust is built. That's what everyone thinks. Like, if I tell you something, I expect that you believe it. But that's not actually how trust is built. Trust is built by that word I said earlier, agency. People need to feel like they are part of the process, they're part of the system. Like I know this from the start. We were together. You showed me this. This is how it works. This is where the pitfalls are. This is how we maneuvered them. Step by step, throughout the entire process, that's how people feel like they have agency. They feel like they're part of the process, they're part of the system. And now this is something that's now ingrained in them. It's like, yeah, I feel like I know this thing kind of inside out. So now I have some sort of level of control. And if I don't know it, I trust that if I call Stella or Melody about it, they'll pick up the phone and they'll take the time to explain this to me. They'll not just send me an FAQ or direct me to a department that is still figuring it out. So the the need for relationship management. That's that's that through change. I say this because I I like business relationship management and a partnerships person. I'm a people person. So what I'm hearing from you, AI transformation needs to have a relationship. There needs to be that how we're deployed. Do you have a relationship with those individuals? Have you created that relationship? Have you cemented it? You've converted them into partners. Am I making sense? Yes, you are. I feel like I'm testing my knowledge with you. No, you are making sense because aside even from like, you know, after the procurement and the people signing off, and even the people using, like, I feel like, especially at this time, yeah, a lot of like the customer service has been outsourced through a lot of chatbots and a lot of whatever, which is a very big part of not even just employee management, customer management, it's a big part because most times I want to talk to a human being. I know. Most times I don't want to hear, I don't know who's telling me now press. No, can I speak to somebody? Like you're not understanding me. So the same thing, like in an organization, I don't want to be using a tool and then you're using and then you're told, oh now go. Please, I want to understand, like let there be like a dedicated team. Yeah. To like, especially if it's like very new, let there be a dedicated team to be there to walk you through these pitfalls, these two issues, so that you're able to go on the other side like successfully. Because again, if it's a new thing and there's like a small barrier, the it slows down the adoption much faster if it was something that was already there. So if it's new and there's just kiddog of friction, the possibility of me like just rolling fully back on it is is very high because I'm like, I'm not involved. I'm not about that life. Let me go back to what I'm not about, you know. But where are leaders getting it wrong with their communication? Because I feel communication is also hit and miss. Where are they getting it wrong? Or where are they getting it right? Or what have you seen thinking I like the way they communicated about AI adoption here? Let's adopt it. Mostly what I've seen that's been gotten wrong is like transparency. Transparency in the sense that yes, we have this tool, but there's these things we don't know, and we are still navigating and we're building. It's great here, but this could happen and it can be expected, that maybe then be outtages. We are still working out a lot of things, and we want to assure you that we are walking this journey with you. Yeah. In case of anything, we're here. Because you they'll just wake up and say, There's this amazing tool. We have Castomakia, we have Engineer Venue, whatever happens, send your issues there, there, there, there, there, there, there. And it's sold like a problem solver. It's solved, it's sold like a one-touch problem thing. Yet we know there's a lot that's still being worked out in this AI space. So just being fully transparent with your people and saying, yes, this is what we have. Okay. However, there's this side to it as well. And this could happen, and these are the guardrails we are put in place to ensure you don't fall into that trap. Exactly. Okay. Okay. And more on transparency. How much should an organization really be sharing with their stuff? Because you also can't be overly transparent. You can't be overly transparent. Like you see how how we've said here, we've said it very corporately. Yeah. Like, you know, there's guardrails, there's this and that. You're not going to go into the intricacies of, you know, so we have this algorithm, and then so the you're not going to be oversharing. Yeah. Because I think a lot of companies are now also putting together their AI policies and stuff and stuff. So they should be like some sort of guardrails, but just not transparency in terms of the details and the models, because again, you'll also be losing people. Not everyone listens to that language. Actually, it's a very few percentage of people that listen to that language. So just in a way that they truly that they are able to understand generally that this is here. There's good, there's bad. But through it all, we are walking the journey with you, and we're here to, you know, protect you, your data and the like. So for striking a balance. Know your audience because there's there's a role that you cannot tell X, Y, Z, because they'll go and translate it. Lost in translation next thing you know that is backed up. Yeah, absolutely. Okay.

Why Social Scientists Belong In AI

Melody Mukhwana 49:52

Now we've talked corporate, trust, responsible AI. I want to go back to something you had mentioned earlier. We need more like you, the social scientists in this field. What would you like to see, especially for somebody listening to this and they study psychology? What would you tell them? Or international relations like me, let me just insert myself down. Somebody who's wondering, what is AI? Nah, AI. How am I fitting into this? Because you now in the larger scheme of things, countries are no longer not trying not to finance the social sciences because it looks like STEM is the winner. But in a few years' time, where did the psychologists go? In fact, I'll now be practicing and making big bank.

Stella Gichuki 50:39

Yes, yes, yes, yes. So before you practice and make big bank, somebody who's wondering, what would you say to them?

Melody Mukhwana 50:47

Wondering how the how. Tell them with AI. In the same way, there's you know that fear of loss, but there's so much opportunity because personally I should not have been in this space.

Stella Gichuki 51:00

Yeah.

Melody Mukhwana 51:00

But God put you there. But I God put me there. God gave me the wisdom to think outside of the box and think, hmm, what can I do different with the knowledge that I have? And that's the same exact thing that's that's coming out around all these things. Like, you know, there's people focusing on ethics, policy, specific things to specific things that people have studied that are not really technical. So there's so much opportunity.

Stella Gichuki 51:29

Yeah.

Melody Mukhwana 51:30

You just need to know what you love. And whatever you're studying, how does it because we need a lot of us. We need honestly, there's we need so much a lot of the human touch because there's big there's a big disconnect when it comes to just the human interaction, the human touch in these tools. There's a quite a big disconnect and it's growing. So there is space for you. Don't don't be afraid to come and take up space. There's a lot of space. You might have to push yourself in there, but for sure. Yeah. The space is there, and we need your voices. We you just need to be curious and find an angle. Yeah. Because I know it's barely explored. There's still so much coming up with AI. Okay. And you can truly find anywhere to fit in. So what you're talking about is coming up and the roles that are coming up, I'm assuming like a responsible AI specialist, an AI ethics resource, to name a few. Absolutely. A people readiness specialists. Literally, that could be a whole new role. Because it needs to be a whole new role. Someone hire me. So you matter, psychologists, you matter. You really, really matter. Don't just take it as, you know, you need to be talking to people only. And you know, even like in this age of AI where people are kind of disconnecting and talking more to their AI than people, there'll come a time where there'll people will be paid for people to talk to people. Listen. And not like like as a psychology, but just like let's talk. That's where the money is. I think it's going to come a time like that.

Stella Gichuki 53:11

And that's what I wanted to close up

Guardrails And Accountability

Stella Gichuki 53:14

with.

Melody Mukhwana 53:14

Actually, a few more questions because the people part, we've seen the harm that this AI is doing, especially in the global north. Yeah. Where somebody, a young person, is being persuaded to cause all manner of harm because of a chatbot or an agency. Yeah. Or try and infiltrate. Was it Windsor Castle that this individual wanted to infiltrate? We're seeing more and more of that. My AI said. My AI said. You know, even workouts. This is I'm following an A. I was there once. Not pointing. Not pointing. But but it's okay. I I like it. And you know, some of these things are really, you know, spot on. Some of it. If you just put your height, your weight, you get a pretty tailored talking about myself. But that's on that's on the positive side. That's on it. Yes. But on this side where your AI is convincing you, you know, I something happened and I and I just joked. I said, my AI said. And it was like, yo. Who have I become? Who have I become? I was joking, but I took a step back. How are we how are we gonna deal with that? My AI told me. My AI said. That's a tough one because and we are still very like young in the in the AI space. And it's having such a huge impact, especially because like you know, our youngest generations are like tech native, they even know better than us. You know, and they are more what's the word, impressionable. Yeah, they're very impressionable, they can follow anything. They will use AI to cre depict people dying or kill them. I know. That's now the harmful part. It's it's really bad. It gets really bad, and I think that's where like the guardrails come in. Because you see, some of these tools, there are some that won't get that far. Okay. There are some that won't allow exhales. There are some that won't allow for that to happen because for they've been made with very tight guardrails that great some because a lot of these tools, as I said, they're they're here for profit. They don't care if you're going to do whatever you're going to do. Uh you want an answer, I'm going to give you an answer. I don't care. I I don't care what you're going to do with them. We've obtained a useful. But there are some that, you know, there's truly, they truly have like, you know, ethics built in it and the like. Which, again, we also want, like, you know, our African innovators to come up with such things that have ethics built in them. They have tight guardrails, they are for us that you know we don't necessarily always. I know it's it's it's expensive, it's a whole other different conversation. Yes, it is. That we do not want to get into right now. Yeah. But at the end of the day, it comes back because that's not something you remedy after the tool is out. No. It's something that's out, it's out. Once it's out, it's out. Yeah. Once you start building, you need to keep that in mind and you need to rebuild with that. Yeah, from the business understanding phase right to through to development and deployment. So you're saying we should build AI ethically. So people like named you. Yes. When we're talking to clients and partners, we have to have our ethical hat on. Because you can't discipline the AI. It's out. It's the individual. So then should governments bring go around and say, who here is responsible? You remember when data the data protection regulations were out in the act? All companies had to have a data protection officer. Do you see that happening with AI? Well, now everybody will have to have an accountable resource. I think so for AI system. And I think it should happen because accountability is so important. Without accountability, then I mean, who's ever gonna know? We can do whatever we want to do. Who's who who's the snitch here to say? So there has to be someone who, in case of anything, any liability, any anything, they'll know it'll come back to them. And so they will do their best and try to make sure that everything is operating within the stipulated guardrails. Wow. And AI AI commissioner? Anyway, anyway. I think there was even there's an office that's been that's a conversation for another day.

Stella Gichuki 57:38

That's a conversation for another day.

Final Message For Leaders

Melody Mukhwana 57:41

Last words from you, Melody. I've really enjoyed this conversation. What was your message to our listeners who are CIs, the C-suite, the CIO, the CTO? Do we need a chief ethics officer now for AI? We do need a chief ethics officer. We need someone who's going to be accountable as you roll out those things. There needs to be someone who's accountable and there needs to be someone who's taking care of your people. Because at the end of the day, if you're just chasing your ROI, you're going to end up losing more in terms of that ROI and your people, because your people, you lose them along the line if they feel not heard, not listened to, not valued, and being forced to do things that they don't understand the value of. And bring them along is like it's unnegotiable at this point. If you want your adoption to move forward, it's unnegotiable to have your people at the center. It doesn't need to be you, but it needs to be someone who you trust and see can carry this forward, is the bridge between you and them. Yes. So the CTO and the CIO need to spread that risk. Yes. Else it will be them. You're the one in charge of tech. No, that's not me. Well, you're the only one here. Why didn't you advocate for some time to help? Is actually happening because when the adoption fails, everyone's like, but because you, who was rolling? I'm like, oh no, it's not me. Now me the CTO or the CIO. The CTO and the CIO are throwing the blame back, and no one knows who's supposed to solve this problem because it's not native to any of them. Whoa. It's kind of like a niche problem. Who's and they're like, well, me, but my tech is working. It's not me. I'm not involved. Yeah. So we're going back to school. And we'll be up skilling. Yes, absolutely. You cannot be sitting in the house doing nothing. There's all the knowledge you have you need to have everywhere. So get a skill or two. Doesn't have to be Coding. Please, there's enough of those. Do do something else, something that you like. And you know what? You can you always have space in this. AI is just starting. There's just so many roles, so much opportunities. So just show up. Show up. Thank you, Melody. Thank you for having me. Yeah, you've heard it from us, guys. Show up.

Stella Gichuki 59:59

There's always a role for you. Responsible AI, people readiness. Um, yeah.