Andreas Welsch: 0:21

Today we'll talk about upskilling your product team on generative AI. And who better to talk to about in somebody that's really, passionate about that. Supreet Kaur. Hi Supreet. Thank you so much for joining.

Supreet Kaur: 0:33

Hey, I'm excited to be here. I think this is off topic of interest to many and for me as well. So thank you for inviting me.

Andreas Welsch: 0:41

Fantastic. Hey, why don't you tell our audience a little bit about yourself, who you are and what you do?

Supreet Kaur: 0:47

So a little bit of my background. In 2017, I decided to come to the States to pursue my master's in data science and analytics. And that's where my journey started on this amazing realm, I would say. And I start my career as a data science consultant. I was working in the healthcare space really impactful use cases like predicting the red. Disease drugs for a patient per month. And then I switched to more of a product and strategy role and I was close to three years in the consulting before I decided that, okay, this life is not for me and I decided to switch to a product role and that is where I am today.

Andreas Welsch: 1:34

Awesome. So sounds like you've already seen quite a lot in different industries and in different spaces. And I know you're super passionate about the topic and speaking as a professional in this space today. So I'm really excited to have you on. Now for those of you in the audience, if you're just joining the stream, drop a comment in the chat where you're joining us from today. I'm really curious to see how global our audience is again. Supreet, should we play a little game to kick things off? What do you say? Yes, let's do it. Okay, fantastic. This game is called In Your Own Words, and when I hit the buzzer, the wheels will start spinning. When they stop, you'll see a sentence and I'd like you to answer with the first thing that comes to mind and why, in your own words. To make a little more interesting. You only have 60 seconds to complete your answer. So are you ready for what's the buzz? Yes. Fantastic. Then let's get started. If ai, where is season, a season of the year, what would it be and why?

Supreet Kaur: 2:39

Ooh. So we are in August right now, so I would say it'll be fall. So if you are based out of East Coast, the weather is. Hot and cold. Early mornings are colder. And then it gets really hot. And that is how the AI hype has been. People have been excited about certain topics and then that hype goes away and people start getting excited about something else. So I would say it's the fall season.

Andreas Welsch: 3:10

Thank you for sharing. Let's talk about the main topics of our episode. And I'm sure we'll, touch on, some of these things as well: how it's super hot, how it might cool down, how it gets hot again. Where do you see the opportunities for generative AI? At the moment for product teams, why is it so hot? What should they focus on and why does it matter to begin with?

Supreet Kaur: 3:34

So believe it or not, AI is going to impact the day-to-day of everyone. It's not just for product teams. I feel everyone is going to get impacted in some way or the other. You might think that the impact is positive or negative but it's coming. And it's here. So I think one thing that I have seen happening is that people are scared. They feel like it's gonna replace jobs but I feel more than replacing jobs. It's going to reform jobs. It's going to give time to professionals to have the creative juices on and think about things that matter more and that require human sentiments, human brain, human connection, and automate things that don't really matter. I think all of us can look around and out of eight hours, I'm sure there are things that we find not so sexy in our day to day. And we don't want to do them anymore. And I think that is where the AI can really serve as a replacement to some of those tasks. And just to name a few writing, meeting notes, creating slides. All of us have been there. All of us have suffered. Where we have writers block, we don't know what to do. So some of these are the examples where AI can really help to automate and transform some of the processes that we do every day.

Andreas Welsch: 4:58

I've been in a PM role myself before. And not only meeting minutes, but also what, does the product do? How do we talk about it? How we describe it? And I think to your point, especially if you have more of a technical background, that's where your core expertise is, that's where your strengths are. But writing doesn't come as easy to us as to somebody who's chosen that as their main profession, right? But yet we communicate through information and through text. So great point here how we can make that easier for product teams. I'm taking a quick look at the chat and it's really awesome to, to see from where across the world people are joining us. Whether it's Portland, Oregon, Zambia, Charleston, South Carolina in the US, New Jersey, Tanzania, Australia, Kenya, Columbia, West Palm Beach in the U.S., India. Fantastic. So thank you so much for joining us. If we talk about these opportunities for product teams and, yes, it's text heavy in many ways, and you don't need to be afraid that your job is going to go away. But there are opportunities where you can leverage it to become more efficient, become more effective, do the things you really like and are good at, and get rid of all the other stuff that we need to do in our nine to five. How can product teams can actually learn about generative AI and how they should use it? What do you recommend they do?

Supreet Kaur: 6:22

Yeah. I think the first step is that as a professional, you really know what your pain points are, and there are times in your day-to-day that you have thought to yourself what if someone else could do this for me? We all have thought about it, so I think the first step is to pin down all those possible use cases. I am a data-driven individual. I've worked with data all my career, right? So I think one pain point that I've always had is, okay, how do I know what all this data means? I don't want to go through these data dictionaries. Is there an efficient way? And if you are working in a larger organization, you would know there are multiple SMEs. Sometimes even finding the right SME might take days. So is there one repository or is there a chat interface where I can go to and search all those data elements and what each of these mean? So use cases like this, I think it's on all of the people who are in those teams to pen it down. And then once you have done that, you can talk to the experts, the AI experts, the AI engineers who can really help you solidify your idea that if it's even possible to be a use case. But I think one thing I would highly recommend is don't be afraid of thinking about all those use cases.

Andreas Welsch: 7:44

I think that's already a very important point, right? A lot of times if there's a new technology and certainly so much buzz with all, these different camps, right? It's, going to help us, it's going to replace us. It's going to this, it's going to do that. It's as revolutionary as fire, but yet will it wipe out humanity? There's so many polarizing points of view. And I feel like it's hard to discern what is actually real? What does it really mean for my role and what can I do? So I was wondering maybe do you have some concrete examples of how can product managers, for example, stay sane in all of this flood of information and dissecting what actually matters? What is nice to know? What's important to know? What's maybe just, for the headlines and the clicks? What's your recommendation? Where should product managers, for example, start?

Supreet Kaur: 8:43

Yeah. I think there is no one way. And I can just tell you how I stay updated is by tuning into podcasts like these. First of all, just to know, okay, what's happening, and then adopt things that are relevant to me. Just because it is new and it's shiny doesn't mean it's relevant to me. So you have to go with that mindset. So I think going with that filtration mindset is very important so that you can choose and absorb the right information. There is a lot of information in terms of articles. There are meetups that are happening. There are conferences that are happening where you can learn about everything, but you don't have to. That's the point. You don't have to. Once you have identified those use cases, then you can talk to other people in the industry and see what they are doing to solve that use case. That is exactly when you'll be able to see the comparison. And I think the other thing, what you really need to do is also think about the return on investment. Just because you have a use case doesn't mean that it needs to be implemented and it'll be cheap to implement. These things are expensive if plan to implement it. And especially if you have been in bigger organizations. If you're uprooting legacy systems and implementing something else, be ready for the pushback and the consequences, right? So you need to think about those things. And I think all of those things come naturally to a PM. I'm a little biased towards them, obviously. So it, does come naturally to us thinking about the business. Thinking about the cost as well as cutting out that noise. And just reading things that are relevant to you.

Andreas Welsch: 10:26

Thanks for sharing. And, to your point attending conferences or maybe even webinars, I think there, there's so many opportunities again now, and especially that things have been open for a while again after the pandemic to really meet in person, to engage, to have that dialogue. And in addition to getting the information here have that exchange with your peers to, to understand exactly what are people already knowing? What what are they exploring and so on? So Supreet, maybe we can take a look at the chat. One attendee is asking, do you know of any LLMs that are particularly good in the field of product management? Maybe we can connect that also with the second question which is, with the emergence of AI powered code completion tools, to what extent will project management governance still be required in the future of software development? And how can project managers effectively integrate and leverage these tools to ensure project success and maintain governance standards? So are you aware of any LLMs for product management and how do you see this evolve around code completion and the collaboration between PMs and software developers?

Supreet Kaur: 11:34

Yeah, I would say that I think obviously it'll depend on your use case. And one thing that I feel is the struggle that there are all of these open source codes and some have. A lot more parameters than the others. So it becomes difficult when you have to demystify them for your use case specifically if you don't have that talent. And I do advise startups and talk to startups who are struggling with the same problem that they have all of these options, but which one to choose. So I think it's a game of exploration and exploitation. I have personally explored the open source code by Meta, which I felt was really good just because it was also easy to interpret and less parameters. It has some great use cases, so I would highly recommend that.

Andreas Welsch: 12:26

How can we make it easier for people that are just starting to learn about this or, that want or need to learn about this new kind of technology? How can we make it easier for them to understand what it is to manage that transition?

Supreet Kaur: 12:39

Yeah, I think it's a two step process according to me. One is that the people who really understand this technology need to hold forums. It could be within your company, it could be outside to explain people who need that layman explanation. And you have to do that same with your senior leadership. They would want to know what you're exactly doing. And the second thing is that you, if you're coming from a business background, need to be ready for that upskill, right? That you need to be ready to absorb that knowledge. And I think that is where sometimes a gap comes in. If you're from business, you might say, okay, I don't have to learn all this. But you do have to. You do have to upskill yourself a little bit to be more comfortable with those kind of things. So if we have that, I would say mutual partnership between the technical teams and the business teams we might be able to achieve the goals.

Andreas Welsch: 13:33

Wonderful. Thanks. I think that puts it very nicely into perspective, right? It's not just the data team's responsibility, it's not just the technology team's responsibility. And it's not just the product team's responsibility. But there's an equal accountability on the business side as well to learn about this to keep up and to understand where can we actually use this? So in partnership, you can identify these opportunities. Great way to, to summarize that. I see there's, another question in the chat around what are the elements that need to be an AI ethics policy, and we've talked about that previously with Reid Blackman and, with Ravit Dotan, on this show. So I would recommend take a look at those two recordings to learn more about AI ethics. I feel we, keep seeing so many articles and research already also that show how large language models are, built in data. From predominantly Western or American parts of the web. And so they inadvertently include and regurgitate those Western values. And I think that's a problem everywhere else in the world where English isn't the dominant language where we don't have Western or even American values. Where's the sense of, well, are we imposing that consciously, subconsciously on other parts of the world? I'm wondering what do you think is required to make AI products more inclusive? And maybe that also goes beyond just the values and regional perspectives

Supreet Kaur: 15:00

Yes, I think that's a great question. We all need to think about. And at the end of the day, even when I'm using all of the AI tools, and specifically when I'm trying to create my headshot or a picture, it's not able to even capture features for a person like me who was born in India and has certain specific features. So I think that kind of validates the fact that you said, that our products are still not inclusive. But I think as they say, AI is as good as your data. There is a strong need to have a detailed analysis as to how does my data look like? Is it representative of everything and everyone in the world, or is it biased in a sense that it just captures Western values? And I feel like there needs to be that kind of accountability in terms of specific teams who take care of that. Just because if you're an AI engineer you're already too close to the problem, if you're working on that, right? So you need that outside perspective just to see if my model is fair enough, right? And is it inclusive enough? And does it accommodate everyone? And the second part I would say of the solution is that the people who are making the decisions, and we all know that when it comes to AI solutions, there is a lot of business involvement as well. Are those inclusive too? What do those people look like? So you have to look around and see, is my room colorful enough? Does it have all the perspectives? Just so that unique perspective can plug in when you are thinking about your product and AI strategy.

Andreas Welsch: 16:44

I love how you phrased that, right? Is my room colorful enough? Like I've never heard that as somebody express it that way, but I think that captures it so nicely in all the different aspects and, facets then that are needed to bring these diverse perspectives and different perspectives around the table and make sure that you cover it from all possible angles. Now, we've talked about regional specifics. What do you think? Are we all going to create content that looks and sounds American and maybe includes references to American pop culture? What risk do you see, if we're using more and more AI to generate content and do we all sound the same at some point? Yeah.

Supreet Kaur: 17:29

I think that's happening already. If you see some of the LinkedIn posts, like they all look the same. They have those very typical icons and, the way that a post is. I think about from a lens that, okay, will I ever talk like this? English is not my first language, right? But even if it was, will I ever talk like this? How GPT presents the views sometimes. And when you see that content and if you think that it cannot be you. Then that's not the content for you, right? You can get inspiration, but you can't really imitate that content and post it, online because ultimately people will know. I use GPT as well, but I use it more to generate ideas and to overcome my writer's block. But I think those are the things that you need to ask yourself and you need to be your own principal in some way when it comes to those things.

Andreas Welsch: 18:25

Perfect. I think very much alike, right? Some of the content that you see looks pretty generic or the other day I was seeing some examples on online where somebody just copied and pasted things and it was like, I'm a large language model trained on data up until September 2021. Pretty obvious where that comes from. Yeah. So that leads to whole different problems and challenges in credibility.

Supreet Kaur: 18:51

Exactly. Yeah. Yeah. You mentioned it, right? Credibility, which is very important.

Andreas Welsch: 18:56

Yes, exactly. So I'm, taking a look at the chat again and there are a few more questions. For example can generative AI enhance the capabilities of product teams in developing conversational AI solutions?

Supreet Kaur: 19:10

Yeah I feel like conversational AI is obviously being impacted. I think with the launch of GPT. Once it was available, everyone started thinking about and having a close look at their existing conversational AI solutions and how they could be enhanced with the generative AI. But what it would mean for product teams specifically, it would impact their day to day. And as I was saying, that some things like building a data bot. To have conversations on the data elements. That is one thing that PMs and others can get impacted. Others could be giving the commands to create slides, right? Or you are managing your Jira tickets day in and day out. All of those things can be reformed and transformed using conversational layer. So definitely I think it would have an impact. I haven't seen any use case coming out where companies are like, oh, we have implemented that. But I definitely feel there will be an impact.

Andreas Welsch: 20:10

So now I know you are in a product role, right? And we've been talking quite a bit about generative AI and the opportunities and how we can upskill. But I'm wondering what excites you the most about generative AI and what's the biggest learning opportunity that you see?

Supreet Kaur: 20:28

I would say there have been people who have been in their roles for 20, 25 years. They've been doing the same thing day in and day out. And when I see their excitement for this generative AI actually makes me excited. They have this very strong background. The subject matter expertise on these particular topics that no one has. Because they've been in the industry for such a long time. That leveraged with a tool that can do all of their I would say mundane tasks. That's a perfect combination right there. And I think that is exactly how we have to imagine our future which is obviously you need to be good at whatever you do, you have to be an expert at whatever field you are in, whatever job you are doing. And then keep thinking about what mundane task can I automate through AI? And I think everyone's gonna be a AI product manager. And you start thinking about that, right? So those are the things that we need to think about and when you will be passionate about that use case. And I, think I do the same when I'm really passionate about, okay, this is what I need to learn. I start talking to other people, I start researching it on internet as to what is happening for that particular use case. I think that is the best way we all can upskill ourselves and I'm excited about.

Andreas Welsch: 21:52

How can anybody not want to become a product manager after that answer. That's, awesome. I'm so glad seeing your excitement and sharing that excitement here with the community as well and what you're seeing in the product management space as well. Supreet, I was wondering as we're getting close to time, if you can summarize the three key takeaways for our audience today from our conversation.

Supreet Kaur: 22:16

I think the first key takeaway is don't be scared of AI. I, know that many of you text me and message me on LinkedIn you know that you're feeling sad or depressed and are really worried about this. I would say you don't need to be any of that. It's not a monster. It's gonna help us. So, look at it with an optimistic lens. And this is the best time to upskill yourself as all of these free resources are available for you. Second, if you really want to start in this space, I would highly recommend Google's certification on generative AI. You can start looking into that. It's free and they start from basics and then they I would say upskill you to a lot of the components of generative AI. The third thing is start noting down the use cases today. You might not know what can be automated in the next year or so. So start thinking like an AI product manager.

Andreas Welsch: 23:17

Thank you so much for sharing that. Look, we're getting close to the end of the show. Thanks for joining us, Supreet, and for sharing your expertise with us.

Supreet Kaur: 23:25

Of course.

Andreas Welsch: 23:26

And thank you to those of you in the audience for learning with us.