39 Comments
User's avatar
Holly Jones's avatar

This made me laugh out loud 🥲 “PRD > (a beautifully formatted 12-page novella admired by many but read closely by precisely zero)”

Expand full comment
Kinga Magyar's avatar

Hi Julie, very provocative article indeed. :)

I was wondering if it really is AI that revolutionizes how companies build products or rather adopting a product mindset (focus on value creation, than on churning out features).

The traditional product development best practice you mention at the beginning of the article (idea -> mocks -> PRD -> code -> launch -> surprise) is already missing important aspects of building a valuable product. Where is the problem discovery? Are we building something that actually solves a user problem and brings us closer to achieving our targeted business outcome? Where are we quickly validating, in this process, whether we are solving an important opportunity for our users and then validating that the solution we ideated actually solves that user problem?

If we sprinkle AI on our product development process, then yes, we might do certain activities quicker, but it doesn’t mean we will do them better. For instance, AI enables people to write a PRD much faster, but it doesn’t mean that results in better product development. It doesn’t mean anyone will read that, not even its author.

I agree that AI enables teams to prototype quicker and run more experiments faster, but in companies where this wasn’t part of the culture until now, AI won’t magically fix that.

I think AI is not the silver bullet. A mindset change is needed to build more valuable products, to focus on the value we create, rather than the number of features. That’s why I don’t think encouraging leaders to “Track speed of feature development: ensure that it’s going up” is the right way. Since business leaders already do this and it results in a lot of waste.

I’d be happy to hear your opinion on this. Thank you for sharing your thoughts!

Expand full comment
Julie Zhuo's avatar

Really good point which I think is evergreen -- the importance of identifying and solving or value continues to be true in today's world and whatever future world of increasingly better tools we have.

Expand full comment
Julie Zhuo's avatar

Really good point which I think is evergreen -- the importance of identifying and solving or value continues to be true in today's world and whatever future world of increasingly better tools we have.

Expand full comment
Joy's avatar

Exactly this. This is the first time I strongly disagree with one of Julie’s articles. Churning out more features isn’t the way to go - data informed strategy is. Strange that she hasn’t replied to your comment.

Expand full comment
Paulo Cassin's avatar

+1 Good point

I think if both are combined: Mindset change + AI throughout the flow, then we extract a lot of benefits and a more streamlined, optimized process for faster value creation.

Expand full comment
Bob Baxley's avatar

Interesting and provocative article but the logic sort of fell apart in your description of the characteristics of good teams.

In that section you mentioned the blurring of the lines, describing how an engineer could produce mockups and a designer could produce marketing copy. But then you also highlight the importance of taste.

My question is, why do you think an engineer has the needed taste to judge the mockups or the designer has the taste to evaluate marketing copy?

The concept of specialization isn’t merely who can produce what. It’s also, and more importantly, about who has spent enough time studying and thinking about a particular field to have acquired the necessary level of taste.

Expand full comment
Julie Zhuo's avatar

Hi Bob! Great points. I do believe that more engineers have good taste in design than could design things themselves, so this evolution of working benefits them. Of course, I also agree with you that many engineers will not have automatically good taste in design, and that time invested is a necessary prerequisite. I do believe it is easier to train the eye than the hand though. And at high enough levels, one can spot many patterns across disciplines.

Expand full comment
KimSia Sim's avatar

I agree with all of these

The ideas will probably catch on once a suitable term that encapsulates these ideas emerge

Too bad “product engineer” has been used

The individual contributor empowered by the age of AI highly resembles a product engineer

Expand full comment
Julie Zhuo's avatar

Personally I like "product builder."

Expand full comment
Arturo Kast's avatar

Good points! I like Product Builder, but I think it's missing the strategic part. To me Product Architect encapsulates more the idea of thinking about product as a whole, but it's too bad that IT has already taken it.

Expand full comment
KimSia Sim's avatar

Hopefully that catches on

Maybe I’m being too engineer I feel that is not precise enough

On the bigger picture, I am agnostic

We just need a better term that catches fire

Expand full comment
Juan Rafael Lopez's avatar

Very interesting perspective. I also think some of this has been missing for years now. Marty Cagan has been pushing for protoyping every day, for along time now. https://www.svpg.com/prototype-of-the-day/ Requirements always just go die on someone's desk or inbox.

I agree this will let us move much faster. I also think we will see sooo many new companies because of it. https://medium.com/p/dbf36c306212 However, many won't last and that's ok.

Expand full comment
Milan Mijatovic's avatar

I feel like I've been in an "existential rut" after a layoff last year. There was a lot here that resonated with me, but seeing your comment "experiment, experiment, experiment" with AI as a constant collaborator truly spoke to me. I felt like I have been doing so many things to try to "level up" but also just for fun and curiosity because sometimes the most interesting challenges are the ones you create for yourself.

I feel a bit more energized after reading this so thank you so much for sharing your thoughts and both opening my eyes but also validating a few things that have been in the back of my mind.

Expand full comment
Paulo Cassin's avatar

On spot! There are to many people fearing instead of embracing AI, but when you try and see that is posible to go from an idea to figma design and then to live running application in a matter of minutes/hours. It's jaw-dropping

Expand full comment
Quentin Mackey's avatar

I love this provocative perspective on how AI will change product management. Instead of generalists I feel the need for specialists is more prevalent now with AI, so the right questions get asked and ideas curated. When I say 'specialists' I mean industry specialists. Gone are the days of just building 'one size fit all' features, customer want solutions that drive outcomes and business value within their domains. I do agree with the reader here who pointed out that the introduction of AI is not just a technology adoption it is equally a mindset (and skill) shift that needs to happen.

Expand full comment
saisameer gadi's avatar

AI can bring in a garden of ideas, Whether it solves a problem for a consumer or customer matters a lot. but even than how relevant they are, how original and creative they are matters. people have better ability to differentiate than a classification model does at the end😁

Expand full comment
Jeffrey's avatar

As long as the rules are being rewritten, let’s remove alcohol from the celebration necessities. As a bonus, it’ll free up cash for more pizza :)

Expand full comment
Julie Zhuo's avatar

Funnily enough I don't drink alcohol anymore so I agree! But when I think of launch celebrations of my past, somehow beer always made it in.

Expand full comment
Akshan Ish's avatar

I’m curious what this does to ICs who want to move into management. Typically they would get to manage a small team first before moving up to director / head of etc. What do you think that path will look like? How will we train leaders? How would management practices need to evolve? I feel like there’s potential for Making of a Manager part 2 :)

Expand full comment
Julie Zhuo's avatar

I think manager-as-coach can continue to have significant influence if the person is adept at it, and there will naturally still be leadership opportunities even across smaller pods. Coaching is also something AI can help us all level up!

Expand full comment
Serena Mariani's avatar

Super interested on your last point, would love to know where do you see the AI for coaching point going

Expand full comment
Rafa Páez's avatar

High agency individuals will thrive. And those who can adapt quickly too.

Yes! Embrace the chaos. The future isn’t built on plans—it’s grown in gardens.

Expand full comment
Russ Wilson's avatar

Hi Julie, AI certainly changes the game. One suggestion -- the 3 questions for pruning -- customer and business value are missing. If it can be built, if it meets users' expectations (if you know what those are), and if it's performant, are all important. But if it doesn't generate value for the business, and if it doesn't deliver unique or better value for customers, it's not going to go far. Thanks for your writing -- always enjoy it!

Expand full comment
Jessica Toussaint's avatar

This article nails so it. I’d update one piece - we need strategic thinkers as those doers. Product managers who thrive in execution but not vision and strategy are going not going to be as effective. Also, I have posted on LinkedIn about how product managers need to learn prompt engineering - this article just reinforces that for me.

Expand full comment
Julie Zhuo's avatar

I still believe execution will matter, it's just that "execution" looks different for PMs -- perhaps it's pulling images to create a marketing page, perhaps it's getting better data that speeds up decision making. There is still some "work to be done!" Yes I do think everyone benefits from getting better at prompt engineering :)

Expand full comment
Alin's avatar

I think a new challenge in this setup could be the alignment / consistency at scale. Take 20 teams moving/learning 50% faster than before—this would require a new way of alignment/ communication to avoid double work or a different investment in AI to help with this orchestration. Any thoughts on this part?

Expand full comment
This Week in AI's avatar

It’s weird how its the way of the world. Specialists make tools that help generalists win .. and inturn hire specialists.

The doom movies aren’t wrong after all.

Expand full comment