When AI becomes more deeply integrated into product development, how will it reshape the traditional product team structure and roles? Until now, product teams have typically consisted of Product Management, Design, and Engineering - roles that remained distinct due to their unique combination of taste, experience, and technical expertise. But with AI models as powerful as o3 emerging this year, we're likely to see significant evolution in these roles, and potentially fewer positions in these areas.
To explore how AI might impact the PM-Engineering relationship, I connected with my friend and former coworker Stephen Weber, Director of Engineering at an AI intelligence startup.
Stephen comes from a background where engineering was heavily involved in product discovery's prototyping process. When I asked if he felt threatened by PMs potentially taking over more prototyping work, his response was surprisingly enthusiastic.
"I really love it honestly, because so many times, a company can be building feature after feature or product after product, and not knowing what to focus on. It's great getting ahead of that risk by building prototypes really quickly, or building them all the way to completion can make it so the risk is moved up. That way, we can actual accomplish those things way ahead of time and build the right things instead of just building things all the time."
We both acknowledged that even in this new era of accessible prototyping, generative research remains crucial to ensure prototypes are guided by real customer needs rather than speculation. Stephen emphasized:
"This means you can test more things that also means you can build some really bad things out there and then there's not that longevity of support right so say somebody builds an app right puts it out there everybody tests it and then they're like oh I'm not interested in this project anymore. And it just gets dumped. And then there's a whole bunch of people, I've seen it before, they get really upset. Like, hey, this is my favorite app. I love this. I put so much into it or something like that. I put all my data in there or something like that. And there's no way to get it out. Like that's heartbreaking, right? We always want to consider like the users and having that empathy. You know, it doesn't always have to be there, right? So I think that's a troublesome part, I would say."
While Stephen and I have always maintained a collaborative PM-Engineering relationship, we discussed how this shift could create tension in highly political companies or situations where either the PM or engineer isn't naturally collaborative. The traditionally healthy tension in the classic "product trio" might evolve into something quite different.
The Design-Engineering relationship is also experiencing significant changes. Sahil Lavingia, CEO of Gumroad, openly advocates for hiring designers who code and engineers who design - both with sharp product sense. This shift is gaining momentum; Paul Graham, co-founder of Y Combinator, recently tweeted about a company replacing Figma with Replit:
"I talked to the CEO of a moderately big tech co who said they'd replaced Figma with Replit. This surprised me because I don't even think of them as being in the same business. But he said Replit is so good at generating apps that they just to straight to prototype now."
(You can see a quick demo I did with Replit here.)
This evolution creates opportunities for product-minded individuals across different backgrounds to build great products, with or without traditional PM roles. The specific structure will likely vary based on company culture, leadership, and optimal product development approaches.
A major inflection point will arrive with OpenAI's o3 model release this year. With its impressive 71.7% score on the SWE-Bench Verified evaluation framework (handling 71.7% of 500 test issues), o3 could transform the PM-Engineering dynamic. If tools like Devin, considered a developer AI team mate, are already managing engineering tech debt without o3, the o3 model's impact on both engineering and PM responsibilities could be substantial. Stephen shared his engineering perspective:
"On the development side though, I'm hoping that it does take a lot of the simpler stuff that we have to go through, project setup or something like that, off our plate honestly. I'm hoping that it's more like we are overseeing the project, coordinating different teams or something like that. Spotting edge cases that we get to worry about the things that the biggest risks or the things that, you know, fall off to the wayside, unfortunately, while you're trying to build like just the bare bones of some feature or product.
And we just get to be a little more high level, like where - I mean, we're still going to be interacting with the code, but maybe we're not hands on as much with the code and we're more overseeing the development of the code and reviewing it."
In the meantime, Stephen, myself, and many other professionals are using AI to amplify our current capabilities while preparing for future shifts. When I asked about his current toolset, Stephen highlighted Cursor, which has boosted his productivity by 30%. He uses it for:
Rapid Ideation: Leveraging it as an IDE to quickly build and test prototypes
Code Context & Navigation: Understanding and navigating complex existing code
Writing Tests: Generating Jest tests with minimal prompting, including edge cases and mocks
Interestingly, I know several PMs who are also using Cursor for prototyping. It's not just for developers - even I was surprised to learn about its code comprehension capabilities through conversation and test writing abilities. The tool's strength in test writing is particularly valuable, given that many startups I’ve worked at struggled with test coverage. With AI, we will see more overlap in the tools that PMs and engineers use in product development, creating opportunities for PMs to better understand technical aspects and transition prototypes more seamlessly to development teams.
Preparing for AI's full impact reminds me of advice given to new parents: prepare thoroughly, but accept that nothing will completely ready you for what's ahead. You can watch the full conversation between Stephen and I in the video in this post. Thanks to Stephen for being on my Substack!
Stay curious.
Anne
P.S.: I’m a fractional product leader and consultant, and I’m currently wrapping up a GenAI initiative at a GovTech startup. If you’re looking for a AI savvy product leader who can amplify themselves with AI, I’ll have availability soon.
P.S.S.: The fifth live rendition of Jumpstart Your AI Career on Pearson/O’Reilly will be February 24. Register here.
P.S.S.S.: I also coach PMs how to 10x themselves using AI in the highest leverage ways, while feeling more empowered and less stressed. I also do corporate workshops on this as well. If you’re interested in working together, please reach out.
