Ep 104: Product Management and AI – Insights from a Microsoft Product Manager

  • 19 Sep, 2023
  • 0 Comment

Related Episodes

Video Insights


The world of technology and business is rapidly changing, and so too is the role of product managers. In the era of artificial intelligence (AI), the responsibilities and expectations placed on product managers have undergone a significant transformation. In this article, we will explore the impact of AI on product management and how it affects businesses and decision-makers.

Embracing Vision and Strategy:

Product managers are no longer just responsible for executing projects; they now play a critical role in setting the vision and strategy for their products. With AI becoming an integral part of many applications and services, product managers need to stay one step ahead, constantly thinking about future advancements and anticipating market needs.

Navigating the “Hype” Cycle:

AI has brought about an explosion of new products, but with it comes the challenge of navigating the “hype cycle.” Product managers must carefully gauge the relevance and value of AI features to ensure they align with customer demands. This requires a deep understanding of both the market and users, as well as recognizing the blind spots and unmet needs that may not be readily apparent.

The Power of Perception:

While user feedback is invaluable, product managers are also responsible for exploring the hidden aspects of user experience. This includes understanding user behavior, psychology, and neurodiversity. By delving into multiple disciplines, product managers can gain valuable insights and improve their ability to deliver products that meet user needs beyond what is explicitly expressed.

Keeping Up with Consumer Expectations:

In the age of AI, consumer expectations are higher than ever before. Rapid advancements in technology have led to an increased appetite for innovation and improved user experiences. Product managers must stay informed about the latest AI developments and filter through the vast amount of information available. The goal is to assimilate knowledge efficiently, keeping pace with evolving consumer demands while maintaining a realistic perspective.

Embracing an AI-First Approach:

With the increasing integration of AI in products and services, product managers are adopting an AI-first approach. This mindset involves reimagining user co-creation, strategizing how to incorporate AI seamlessly, and leveraging its capabilities to enhance products and experiences. It is about utilizing AI at the right moment to create meaningful, personalized interactions with users.


The role of product managers has evolved significantly in the age of AI, from executing projects to becoming visionary strategists who lead their products into the future. They must keep up with the breakneck speed of technological advancements, navigate the hype cycle, understand user perception, and meet ever-increasing customer expectations. By embracing an AI-first approach, product managers can harness the power of AI to create products that make a real impact in the lives of users. In this dynamic landscape, businesses must recognize the importance of effective product management and invest in nurturing their product managers’ skills. By doing so, they can stay competitive, deliver innovative solutions, and create meaningful experiences for their customers in the world of AI.

Topics Covered in This Episode

Primary Topic: Product Management and its Evolution
– Explanation of the role of a product manager in setting vision, strategy, and roadmap for the future
– Differentiating product management from project management
– Considering the hype cycle and anticipating future developments in AI and generative AI
– Product managers wearing different hats and exploring unmet needs and blind spots through disciplines like neurodiversity and neuroscience
– Balancing market trends and user feedback with hidden perceptions and unspoken needs
– Challenges of keeping up with consumer demands and unrealistic expectations in the new age of AI

Primary Topic: Changes in the Everyday Life of a Product Manager with AI
– Transition from cloud-first to mobile-first to AI-first approach in product development
– The concept of building products with an AI-first mindset without ignoring non-AI aspects
– Three styles of product managers evolving in the past six months: platform AI product managers, day-to-day working

Podcast Transcript

Jordan Wilson [00:00:17]:

The role of a product manager is changing more than probably ever before. So what does that mean for for all of us using these products in the new way age of AI? That’s one of the things that we’re going to be talking about today on Everyday AI. Welcome. So excited to have you here. My name is Jordan Wilson. And what’s Everyday AI? Well, it’s a daily livestream, podcast, and our free daily newsletter. So make sure you check that out. Helping everyday people like you and me keep up with what’s going on in the world of AI. And there’s a lot. So I’m excited to have our guest today who is a Senior Product Manager for Microsoft. But before we get to that, and I hope you’re excited because I am as well, but let’s first talk about what’s going on in the world of AI news because there’s a lot going on as always.

Daily AI news

So first, there are some big Google Bard updates. So hours ago, Google just unrolled new updates to its large language model, Bard, connecting it to YouTube, Google Drive, Gmail and other products. It also added a fact check button. So I did give it a quick run this morning. Not super impressed yet. Even with the fact check, it kind of brings up a similar Google search, but we’ll let them dust off the bugs and check it out again in a week or so. So make sure to check out more in the newsletter today about that.

Also, Microsoft speaking of, has a big special event coming up in New York City on Thursday. So recent report this morning suggests that not only will Microsoft be potentially unveiling some new Surface models, but also detailing some new AI features that could be rolled out across many of its Windows apps. So again, check out the newsletter for more on that.

Speaking of AI and large language models and changing the way we work, there’s even more AI writers on the way. So just announced writer inc. It’s a startup that uses AI to help businesses write and summarize content. If you haven’t heard of writer inc. They are more geared toward enterprise clients. But they just announced a 100 million dollar fundraise recently. So Writer they says that Writer Inc. says that they can produce written content as if it were written by a human. I don’t know about you all, but if you’ve taken the prime prompt polish our free PPP course, I feel you can do that.

Intro to Rolly Seth and the role of product manager

All right, all right. All joking aside, let’s talk product management. I’m excited for our guest today. I’m excited to talk product management because it’s something we haven’t talked about on the show yet and it’s something that actually affects us in our everyday life. So bringing to the everyday AI show. Please help me welcome Raleigh Seth who is the senior product manager at Microsoft. Raleigh, thank you for joining us.

Rolly Seth [00:03:14]:

Thank you so much. Good morning. Good evening folks around the world.

Jordan Wilson [00:03:18]:

Yeah, it’s true. It’s a very early morning for you because you’re joining us from the west coast. So I have to start giving out special awards on the show for people that join us live at 05:30 a.m. Their time. So thank you for that.

Rolly Seth [00:03:30]:

Thank you so much. It was fun and thanks for giving me the opportunity to join.

Jordan Wilson [00:03:34]:

Oh, absolutely. So, as a reminder, thank you for everyone joining us live. And if you have questions for Raleigh, please get them in. And as a reminder, if you want to know all the secrets of what Microsoft is working on, that’s not what we’re here to talk about today. So if you want to know in general about product management and how the industry is changing, get those questions in. But let’s start at the top rally. Can you actually explain to our audience, because not everyone’s a dork like me, but could you explain what product management is and what product managers in general work on?

Rolly Seth [00:04:07]:

Yeah, product management is a super interesting field that has evolved over the last decade or so. But more or so it is different from project management in terms of product managers are supposed to set the vision. So while you guys are thinking and using chat GPTs, we are already thinking one step ahead where the future is evolving. Say Gartner’s hype cycle came out. What are the hype cycles beyond the generative AI thinking? So we have to think in terms of the next vision, strategy and roadmap building beyond what is currently available.

Jordan Wilson [00:04:39]:

And that’s so interesting because you talk about hype cycle and I think with so many products that have AI inside of it, it seems like it’s hard to gauge that. How does that even work? I guess in the day to day of a product manager, are product managers seeing how everyone’s using their products and looking at bugs reported, looking at features? How does it actually work to what people experience when we sit down and use the products?

Rolly Seth [00:05:11]:

I would say product managers wear different hats. It’s not just one that you just look at the users thing because users are giving the feedback of what we have shipped. So we have to keep an eye on not just the market but also unmet needs. For example, when you talk about PPP, I also have for AI, specifically a PPP of the equation called AI magic equals to plot perception and presentation. But the part that usually project managers don’t cover is perception. And perception I mean, is what users don’t tell you because they don’t know. Maybe that’s a blind spot. So that’s where we explore in terms of, for example, neurodiversity or neuroscience. Sometimes those are even interesting concepts in the AI field that people are not usually but in product manager we want to explore multiple disciplines. So I would say we wear different hats. It’s not just hey, look at the market, where it’s going, but figure out what’s hidden also.

Changing consumer demands with AI

Jordan Wilson [00:06:06]:

That’s fascinating. Tell me this because I’m very curious because I think the product lifecycle has changed dramatically with AI. Probably in building the products, but also probably customer expectations. Because we just assume in this new age of AI where everything moves so quickly, when we get these new products to play with, it seems like every day. How can product managers in general keep up with all of that? Because it seems like consumer demands in this new age of AI are probably unrealistic.

Rolly Seth [00:06:47]:

Yeah, and I would plus one to that. So again, taking the analogy of magic with AI, usually the concept is the more you know, more you will miss. That’s super interesting in terms of information knowledge that you’ll think, hey, if I’m on top of everyday AI, maybe I will grasp more. But in terms of sometimes you have to have multiple diversity. Not just consume all the things, but filter it out. And that’s where product managers have to have an interesting role of hey, just how do you consume different set of information? So I know it might be slightly different concept that you would have heard, but yeah, just assimilating.

Jordan Wilson [00:07:26]:

Hey, as a reminder, thank you everyone for joining us. So Ahmed, the first comment. Yeah, love it, bronwyn. Joining us from South Africa. Brian, thank you for joining us from the Gulf Coast. Rajiv talking about the you appreciate know Raleigh, I want to ask you because you’ve been at Microsoft now for more than a decade, right?

Rolly Seth [00:07:51]:


How the product manager role has changed with AI

Jordan Wilson [00:07:51]:

So how has even just the everyday life of a product manager changed with AI? I don’t know, but I assume that maybe eleven years ago it wasn’t quite at the breakneck speed that it is today. I could be wrong, but how has even the role of a product manager changed over the last decade?

Rolly Seth [00:08:12]:

Yeah, I would say in terms of the entire tech industry, the evolution I’ve seen in the last decade is maybe we used to say cloud first to mobile first and now AI first is one of the approaches that across the tech industry they are taking up how do you build product AI first? That doesn’t mean ignore the non AI part, but how do you start thinking that user co creation would happen with AI? How do you bring in at the right moment AI? So in that terms, I would say there are three styles of product managers evolving over the last six months or so. One is platform AIPM product managers which is hey, we’re just day to day working on latest LLM models, trying to debug, trying to optimize. The second part is AI feature product manager, which I categorize into so they don’t do LLM development but whatever LLMs have been developed, they customize the prompt on that do prompt engineering. So that’s a major bucket. I would say around 40% to 50% of current product management in AI are around that. The third bucket is where industry is evolving is AI adopter PMS and that is basically applied AI that maybe they are not shipping AI features but they want to use for their day to day productivity AI tools. And as per the Harvard Business Review recently, they’ve also mentioned in next couple of years if more product managers start adopting AI technology, their productivity and success rate of the product will increase by 25%. I don’t know if you saw that data. So it’s a huge shift in how product managers can increase the success rate.

Jordan Wilson [00:09:44]:

Yeah, I love how you laid out kind of the history going from kind of a cloud focus, mobile focus and now it’s AI focus. Right. How can product managers keep up with that especially? I feel it’s probably different with a company like Microsoft that has world class access to AI at every point of the product development lifecycle. But for other product managers out there who maybe are tuning in and listening, how can they balance that AI first focus with accessibility to the tools or even with keeping a close hand in hand relationship with their customers and with those using the products? If there is this AI first mentality.

Rolly Seth [00:10:40]:

Yeah, I would say in that terms what AI has done, it has democratized. So it’s not that we have special access to the tools that maybe everyone will not have chad GPT we all use. I would say you guys have the right tool sets in terms of how PMS can stay updated. The one area that everyone I would highly recommend product managers to spend more time is prompt strategy. I’m seeing like prompt engineering people talk about but it is usually because of the engineering work. People say it’s an engineering job, but for prompt strategy I’ve not seen a lot of templates and how people are evolving and by that I mean as part of there is also a cool YouTube video if people are interested by Duolingo AI on how, as part of the product management cycle, prompt validation as far as coming early up before you start even designing your product. So I would say if one thing you have to do for product manager keeping up, try to just explore, keep typing and seeing how prompt different prompt structures are changing personalization are changing, voices, is changing the response.

Tackling the basics of prompting GPT

Jordan Wilson [00:11:46]:

You mentioned something in there just prompt engineering and I think that it’s a skill set that I think everyone’s hearing about but very few people know what it actually means because it seems like everyone’s giving it a different definition and it might be depending on what someone’s trying to sell you on. But just can you give a little bit more advice on how product managers or even anyone should take on this role. I tell people it’s a skill as crucial as searching the Internet, so it’s not as dependent on even what your job is. But you should know how to chain together multiple prompts because it’s the future of work. So what’s your advice to people when tackling the basics of prompting?

Rolly Seth [00:12:34]:

Yeah, usually people start with the basics, like, hey, I have a need, I will just type it out. But there’s the thing of context. I know you’ve talked a lot of time. Context. That’s one of the piecing where prompt engineering or prompt strategy is like the chat. GPT versions can give you quick answers, but how do you give the role and the task? So usually prompt is where you give a task and have a goal that’s like input and output. But where we are saying and product managers are involved in is not what meaning it’s not just the output, but how the output will be generated. Think of creative ways. So how can you manipulate the style tone? Those are very subtle details that people might not think about and might not type in the prompt, but are super interesting. And it’s not just in terms of like resolution or other things, but, hey, it needs to be friendly, it needs to be conversational. All those human values sometimes have to be matching. So I would say just the tone and the style could be some of the aspects that people could add for prompt rather than just input output.

Jordan Wilson [00:13:35]:

Yeah, it is so crucial and I kind of laugh and poke fun sometimes on the show, people sharing their prompts. And I tell people that’s like saying you wanted to learn the Internet by using someone else’s BOOKMARKS. Like, you really have to learn how to prompt on your own because it is an essential skill. So thank you everyone for joining us. Bronwyn talking about she’s learning prompt engineering. Harvey joining us. Michael joining us. Have a question here. Raleigh, I hope you can take this one. So Ben, joining on LinkedIn says, how is the industry using AI to improve or simplify product management?

Rolly Seth [00:14:20]:

Yeah, so I would say it is still evolving in terms of simplifying, it’s more of just not the tools. I would say it’s the process and the mentality. That’s how the industry is evolving. That as I was mentioning earlier, just in terms of the product lifecycle, usually it was, hey, we envision define product, develop iterate on it, but just understanding that now it’s a co creation, that it’s not just developed on your own, that as a human, I’ll tell. But sometimes you have to plug and play based on the need. So what point should we invoke AI? These are like tiny examples. Should it be always on? No, I don’t want AI to be always on. So in terms of how do the industry improve, it’s not about chatbot versus non chatgpt. It’s about where you can start plugging in co creation needs. And based on the industry, it could be different. And based on the confidentiality trust and data requirements, it could be.

Rolly’s favorite AI tools

Jordan Wilson [00:15:15]:

Such a that’s such an important point that you bring up. Oh, and I have to get to the last part of this question too. Everyone always loves asking favorite tools. So, Raleigh, what are your favorite tools to use? I’m guessing he’s talking AI tools. So whether it’s in your role or outside of your role, what are those kind of favorite AI tools that you’re using a lot?

Rolly Seth [00:15:38]:

Yeah, I use even not just Jack GPT everyone uses, but also hugging face. Every new model that is there, you can test out without setting up your own setup and playground. So I’ve been using, for example, for video storytelling, people don’t use Depth Map and there’s a hugging face. So hugging Face is just one tool where I need to just understand an assess. Another one that I’m seeing a lot of potential that we use across is Figma for prototyping, for early envisioning. And Figma has been coming up. If folks are interested, just watch the YouTube recent series on Config 2023. It was done last month, I think, but they had a lot of cool experiences in terms of how core development that I was talking about co creation with AI they are working on and they’ve started releasing plugins, I would say it’s still not mature fully. But in terms of how they’re adding these tiny bits of it’s a product called Genius by Figma that they’re also so those two, I would say, apart from the standard ones I have been exploring and interested more.

The role of storytelling in product management

Jordan Wilson [00:16:47]:

Rarely. We very rarely talk about hugging face on this show, but it’s actually one of the most funded AI companies out there because people, developers, product managers, everyone’s spending so much time on there playing with different open source platforms, which I think is extremely important for the development of software and user experience in general. Thank you. Everyone shouting out the PPP this. How do you when we talk about storytelling, Raleigh, specifically around products, right? It’s something we talked about a little bit before the show, but how can we tell better stories around products? And how do you think that’s changing? Because we have to be able to tell a good story about all the features of a product and then we’re able to use these products as well to just tell better stories. So how does storytelling play into product management?

Rolly Seth [00:17:50]:

Yeah, I would say I did a recent survey on AI product management of how people are using and stakeholder communication is a big piece. They are expecting it would help. And stakeholder communication is becoming more interesting with AI because there’s more and more as I was mentioning information. But people have more and more time with stakeholders and partners. So stories have to be very crisp. And by crisp, I mean I’ve started exploring nano learning. So it’s not even similar to YouTube short, how you can just provide less than a minute information for people, just for stakeholders. So in terms of storytelling, I would say just consuming the knowledge and condensing it for quick consumption. That’s where creatively we are trying to it’s much more beyond, I would say, than saying, hey, use the summarization feature of YouTube or something. But just how do you have the context and nuggets? Figure it out. So I would even give a question there because you’re curious.

Making your product’s story stand out

Jordan Wilson [00:18:47]:

Oh no. Yeah, but that part is so fascinating, being able to use AI to tell better stories. And I love what you said there just like nano learning, right? It’s like a minute long video. I think in some cases people might start to lose their attention. Now it’s almost like generative AI has allowed so many people to become storytellers. So I think it’s even that much harder to get your messaging to stand out. So even like let’s put on the role or on the hat again of a product manager if you’re wanting to tell the story of your product. How can people do that in this age now where it seems like we’re getting way more messaging, way more marketing, way more communication than ever before. So how can you kind of whether it’s using AI or not, but how can you make that story of your product stand out?

Rolly Seth [00:19:47]:

Yeah, one of the examples I would give is like virtual cinematography. Although people say, hey, it’s for film industry, but even product managers can adopt and the example would be storyboarding. So I don’t know if you have heard Jordan, but Leonardo AI, they have started exploring storyboarding AI. And by how that would be, they’ve not started releasing, but they’ve started just showcasing in public media. An example is storyboard. It’s not just quick final video creation, but you have frames, you have scenes. And how do you create scene by scene different frames and have a concept story that you can do within a minute? I’ve never seen because I’ve been part of some of the virtual cinematography earlier. So I would say in that terms, there are certain tools like mid Journey, which will generate single images. And then there are I’ve seen as part of I’m fully updated on AI video creation tools where they’ll teach you use Runway ML and did to generate video. But there’s a midway where product managers get involved. You get to choose of what angle. A lot of all features, I think Pica Labs on Discord released their camera runway ML released their camera function. That’s why because you get to control at what angle, at what speed that story is.

The future of the product manager

Jordan Wilson [00:21:03]:

Know there was a lot there to unpack. If you aren’t able to keep, know, writing your notes fast enough, don’t worry, we’re going to be recapping everything that Raleigh’s sharing about in the daily newsletter. So don’t worry. Stick with us here. But when we look Raleigh into the future, which I know is hard to do, but how do you see the role of the product manager changing? Because we kind of joked about this pre show, but working at Microsoft, I feel that sometimes you all might be working in the future. Right? So what does the future look like for the rest of us and specifically for product management? How has it changed and how do you think it will continue to change?

Rolly Seth [00:21:47]:

Yeah, I would say another quote from Gartner. They have predicted by 2030 that at least 80% of some of the product management project management jobs might be taken not taken over, but complemented by AI. So in that terms, the main angle where product manager can keep focusing as creative thinking and analysis. So, one of the trending areas I’ve seen, again, going back to our initial conversation of hype chart AI simulations, not many people have started to explore of how simulation for future because you’re currently having a need, you just predict for now. But how do you predict it’s not about just diseases, but in tech industry, how do you do simulation? But another one is the trust. So there is a term called AI trism, T-R-I-S-M which is about this trust evolving. And there’s another game if folks are interested in the trust and the privacy part, gandalf AI. How basically it gamifies learning prompt engineering and prompt injection. Another technical jargon. But prompt injection is just understanding how do you figure out and the right way to prompt. So, I would say where the future is evolving. I’m just using all the public references of hey, it’s not just plugins and extensions, although I love a lot of extensions, but there’s something more of trying to use it for different yeah.

Jordan Wilson [00:23:12]:

Wow. So we’ve covered a lot on this show. So if you’re just tuning in now, make sure you hit that rewind. Raleigh just gave us a whole lot to digest. So we talked about everything from kind of the history of product management, how you can use AI in your product management, the power of telling stories. But maybe riley, what’s your one takeaway again? I do think that there’s going to be a lot of product managers listening to this. So with your great decade plus of experience at Microsoft, you’re in the room where a lot of the future of AI is being built and being implemented. But what’s your kind of one takeaway for maybe product managers that are tuning in and who are saying, I want to grow my career, I want to keep up with everything AI. What’s kind of your takeaway advice for everyone.

Rolly Seth [00:24:09]:

Yeah, I would say more than just reading do it. So, my perfect example is because I categorize myself ABC, which is always be curious. And I leave with the quote of when was the last time you did something for the first time and that applies to AI. Also, what was the last AI tool you experimented with even if you didn’t use it? Like day to day? So keep exploring.

Jordan Wilson [00:24:30]:

That’s how you that love. That such great advice. So Rowdy, thank you so much again for joining the Everyday AI show and taking us on really a history lesson of product management. We really appreciate it.

Rolly Seth [00:24:45]:

Thank you so much for inviting. It has been a pleasure.

Jordan Wilson [00:24:48]:

All right. And hey, as a reminder, like I said, we’re going to be recapping it all. So just make sure you go to youreverydayai.com that’s youreverydayai.com sign up for that free daily newsletter. It’s going to be coming out in a couple hours. Raleigh mentioned so many cool tools. A couple of them I haven’t even used, which is crazy because I use just about everything. So don’t worry, we got you check out the newsletter. Thank you again for tuning in and we hope to see you back for another episode of Everyday AI. Thanks, y’all.

Leave a comment

Your email address will not be published. Required fields are marked *