Team dynamics are changing, the tools are changing, and the fundamental roles of designers, product managers, and engineers are changing. In this new world where the only constant is change, what does it mean to lead a team?

This year, LinkedIn is taking a big step towards evolving its entry-level Product roles. 

We created something called the Associate Product Builder Program—and the expectations are completely different. We're now looking for candidates who can do product management, design, and engineering, all using AI tools. One candidate I interviewed had built an entire content creator matching platform from scratch. It was a functioning product with over 100 brands and creators, complete with real users and actual revenue. He used Webflow, Airtable, Wiz, and Figma to build it. Five years ago, finding someone like that would have been impossible. Now, I'm seeing it regularly.

This shift didn't happen in a vacuum. Our CPO made the decision specifically one year ago because the tools have evolved to the point where people can genuinely do all three disciplines with reasonable competency. Five years ago, entry-level candidates showed us typical UX portfolios with the double diamond process based on hypothetical constraints and users, or PM candidates did theoretical business case interviews. The roles were more one-dimensional—you understood other functions and could partner with them, but you weren't expected to actually do their job. Now we are expected to be proficient in stepping in when needed.

As a Senior Design Manager at LinkedIn overseeing the Profile and Pages product, I'm watching traditional roles transition in real time. The challenge isn't just adapting to change—it's providing clarity and conviction to move forward when the path isn't obvious. Here's what I've learned leading a team through this transition.

The pressure to integrate AI into our workflows is enormous. Yes, teams can ship faster now, but speed without discernment is dangerous.

We regularly use Glean to synthesize user research reports, and it's generally good at giving you a high-level overview. But I've been working on LinkedIn Pages product for eight years, and I can tell when important context is missing from the synthesis. The AI is only as good as the reports it pulls from, and it doesn't have the tribal knowledge or historical context that comes from being deep in a product area for years. I still have to look at the sources that it’s pulling from and skim through those sources to ensure that the data is accurate to avoid hallucinations. 

The pattern I'm watching for—both on my team and across the industry—is people using AI to shortcut their thinking entirely. I've seen designers create design principles and product evaluations generated by AI that look polished on the surface but lack any real context about our product ecosystem or members' actual perspectives. The output is generic and ultimately not useful.

This is the trap I notice, AI can make something that looks good and sounds logical in minutes. However, if you're not applying critical thinking or if you're not pressure-testing the output against your deep knowledge of the product and users, you're just shipping faster garbage.

So in design reviews, I push on this. When someone presents AI-generated concepts, I ask questions like: What edge cases did you consider? Where did you start from with the inputs? What's the downside of this approach? How does this connect to what we know about our members from actual research? The AI can help you generate options, but the critical evaluation, that's still entirely on you.

The fundamentals haven't changed. You still need to be good at your craft and have strong design foundations. You still need to understand users deeply and understand good design principles. AI is a tool to enhance that foundation, not replace it.

The pace of change means you can’t afford to be slow. Moving fast doesn’t mean abandoning quality or rigor, it means being strategic about where to invest your time.

Recently, my team was working on a series of comprehensive product audits, and the instinct from the designers was to take more time to be thorough with all the details and cross-referencing sources. However, another team had already started identifying opportunities in the same product area and had ideas ready to propose on the roadmap. If we took another month to perfect our work, we’d miss our window to influence the product direction.

Alternatively, I could have encouraged the team to use AI to quickly generate competitor research and a product analysis. It would have been faster and would have gotten us some basic actionable insights, but it would have resulted in generic and surface-level content, exactly the kind of output that wouldn’t actually move the needle.

Instead, we delivered something in two weeks that was grounded in our deep understanding of member needs, interaction design opportunities, and understanding of the metrics to move. It wasn’t as polished as it could have been with more time, but it was infinitely more valuable than an AI-generated analysis because it had the context and nuance that only comes from really knowing the product and users. We used AI as an editor to check the analysis and generate quick design ideas to illustrate the possibilities. It didn't replace our foundational thinking—we only used it where the risks were low.

This is what conviction looks like in practice. Trusting that your foundational expertise even delivered quickly and imperfectly is more valuable than fast, polished, but shallow output. You move with speed, but you don’t shortcut the parts that actually matter, the critical thinking, the deep user understanding, and the strategic point of view.

AI can help you move faster, but it can’t replace the judgment about what’s worth doing quickly versus what needs more depth.

The Associate Product Builder Program isn't just about entry-level hiring. It's a signal of where expectations are headed for everyone.

I'm seeing designers on my team pick up coding skills and use tools like Cursor to contribute directly to implementation. We're in the middle of a major Server Driven UI migration right now, and like any large-scale engineering effort, teams are under pressure to ship fast and hit milestones. The engineers I work with care deeply about quality, as some of the best engineers I've worked with are meticulous about these details. But when you're prioritizing what needs to get done first, visual polish like spacing, font sizes, and color consistency often gets pushed to later in the process.

This is where designers who can code become incredibly valuable. When we can jump in and handle those details directly—adjusting padding, correcting spacing, updating CSS—we help balance quality and timelines without adding to engineering's workload. As designers, we know the experience better than anyone else, so we're often best positioned to make those refinements.

That's what "full-stack builder" actually means in practice. It's not about replacing other roles, it's about being able to move faster and maintain quality standards without being blocked by other teams' priorities.

The traditional, siloed roles are blurring. Our CPO talks about this concept openly: in the traditional lifecycle, you have a PM thinking about strategy, a designer thinking about designs, and an engineer thinking about implementation. Is there a world where that could be one person, or someone capable of doing all three?

The reality is that expectations are shifting. Just being really good at Figma prototypes isn't going to be enough. Expanding your skillset into adjacent territories makes you more effective and more valuable.

The thing that weighs on me isn't the technology, it's thinking about the broader talent landscape.

I'm concerned about how junior designers break into the industry now. Job prospects are shrinking. The bar for entry is higher. We're expecting candidates to already have experience that they used to gain in entry-level roles.

But what may seem bleak holds an opportunity. The same tools creating this challenge are also creating a solution. If you can't land that entry-level role or internship, you can now use AI tools to build real products from scratch. That candidate I interviewed with the content creator platform? He built that on his own, gained real users, and demonstrated the exact skills we're looking for. Five years ago, you needed a company's resources and a team of designers and engineers to build something like that. Today, you can do it on your own with better tooling and AI assistance.

So if you're struggling to break in, my advice is don't wait for permission. Build something, ship it, and get real users. That's increasingly what will differentiate you.

Across the industry, I see people at all levels navigating this transition in different ways. Some are slower to adopt new tools. Others adopt them quickly, but without the critical thinking that makes them truly effective. Finding the right balance by using AI to enhance your work while maintaining a strong human perspective grounded in context is key.

I don't have perfect answers for these challenges. As leaders, we're supposed to be the stable force with all the solutions. But I think it's important to be honest, I'm figuring this out too as things evolve. The landscape is changing faster than any of us can fully process.

The most resilient leaders aren't the ones who have it all figured out. They're the ones who can admit uncertainty while still providing direction. Who can push their teams to move faster while also creating space for learning and adaptation.

Here's what I'm focusing on now: helping my team build skills incrementally and encouraging them to use AI when it makes sense, celebrating when they step outside their comfort zone, and creating opportunities for them to experiment with new tools in low-risk ways. I'm also being transparent about where the industry is heading and what skills will matter the most, so they can make informed decisions about their own development.

This transition is messy and sometimes uncomfortable, but it's also opening up possibilities that didn't exist before. The designers who lean into learning, who stay curious, who are willing to be beginners again will be the ones who will thrive.

We're all figuring this out together, but we don't have to do it alone. Share what you're learning, ask for help, and try things that might not work. That's how we move forward together in this uncertainty.