The Most Dangerous AI Mistake Product Teams Are Making Right Now

How to escape from building stuff that doesn't matter

Building with AI is a double-edged sword.

You can build faster than ever before, and at the same time, you can get more distracted than you imagine.

I’m worried about what’s happening now. How often do you hear one of the following?

• Which AI features could we build?

• How can we build this faster with AI?

• Let’s only hire software engineers who build with AI.

I hear that several times every day. And that’s scary.

Am I against AI? Not at all.

Am I against building things nobody needs? Yes, 100%.

Years ago, Jeff Patton, author of User Story Mapping, said:

“Your job isn’t to build more software faster: it’s to maximize the outcome and impact you get from what you choose to build.”

This quote still resonates with me a lot; it’s not because you can build something that you should do it. AI does enable you to ship stuff faster, but are you just shipping, or creating value?

Let’s have a serious conversation about it.

Keep reading if you care about creating products that matter.

Stop here if you believe AI can do everything for you.

Critical Thinking Is Your Responsibility, Not AI

Regardless of where you work or what you do, critical thinking is a skill that propels your professional growth. What does that mean?

• You can ponder multiple options and choose the one that makes the most sense

• You can evaluate what works and what doesn’t

• You can distinguish signs from nonsense

If you’re using AI, which I recommend you do, you need to recognize that you’re the driver, not AI.

No matter how fast AI evolves, it’s still your co-pilot because you should be the one calling the shots, not AI.

You should be worried about becoming a prompt master who shifts responsibility to AI and stops thinking about what makes sense.

Sadly, many people are ignoring the basics right now. Yes, AI can help you build faster, but do you know if you should build that in the first place?

Some things don’t change.

You still need to uncover problems worth solving so that you can craft solutions worth shipping. In this order, not the other way around.

How Can You Avoid the AI Feature Trap?

In his book Running Lean, Ash Maurya stated one key aspect of creating digital products, which is especially relevant in today’s world:

“Your learning speed is your unfair advantage.”

Many people ask the wrong question: How fast can you build?

A better question would be: How fast can you drop bad ideas?

We’re fast with ideas, and AI now enables us to transform ideas into solutions. However, this approach overlooks the crucial aspect of building what matters: determining whether our idea can withstand interaction with real users.

It’s not because we have an idea that it makes sense.

First things, first.

Product experiments remain as relevant today as they were before the AI hype.

You need to develop an experimental mindset to enable you to discard flawed ideas quickly.

1. Selecting Ideas Worth Pursuing

We’ve got to slow down our horses. Remember, it’s not about shipping more features, but about maximizing value.

Whenever we have an idea, the first thing we should do is evaluate it. Not all ideas deserve your attention.

I recommend filtering out distractions and then running experiments.

1. Ask how your idea helps you get closer to your vision. Then, discard those unrelated to it.

2. Confront your ideas against your current strategy, and drop those unrelated to it.

Within these two simple filters, you eliminate distracting ideas, and then it’s time to run experiments.

2. Building the Experimentation Muscle

Experiments aren’t about proving yourself right; they’re about accelerating learning.

One aspect people misunderstand. The less you know, the faster your experiments should be. What am I talking about?

A/B testing is often the default method for many teams, yet it takes a lot of time to gather substantial evidence. You can use other methods.

What I like doing is the following:

• Start with experiments that you could gain signs in hours, not days. For example, surveys, landing pages, and other similar tools.

• As you learn more about directions, increase the investment. Run interviews, interactive prototype sessions, and similar methods. We’re talking about a few days, not weeks.

• For promising ideas, go for more robust experiments, such as the Wizard of Oz (fake it until you can make it), a quick & dirty solution for a small group of users, and other similar approaches.

Increase the investment in the experiment as you evolve your understanding of the problem space. You may well use AI to create experiments, prototypes, and test your idea. At this iteration, your objective is to learn, not to ship.

Dropping ideas is as successful as shipping good ones.

You don’t want to bloat your products with irrelevant features.

One trick I have for you is to use tools like Claude.AI, ChatGPT, Gemini, or others to help you explore experiments. Here’s a prompt I continuously use.

Considering this idea, "Your Idea Description" act as an experienced product manager and do the following:

1. Define 15 product experiments to my idea. 
The first 5 experiments should be possible to get results in 4 hours, then 5 experiments in two days, and 5 that you can get results in 4 days.

2. Each experiment includes the test method, sample size, success criteria in absolute numbers, evidence strength (elaborate on the rating), and step-by-step execution

3. Present the results in a table

Consider less conventional test methods.

When using AI, always ask for more options than you need. Then, you can choose the one that resonates the most with you, and iterate to your taste and need.

3. Deciding What to Build and What to Drop

Based on your experiment results, you should learn that some ideas are just noise. Drop them and move on. Celebrate that you learned that fast enough.

The promising ideas deserve your attention, and you should determine how to deliver them effectively. At this moment, building with AI makes total sense.

You can use any of the tools you like, such as Cursor, Lovable, or Bolt.new, V0, and others. You may have used them before to test your ideas, which is more than fine. However, at this moment, you want to ship value because you learned what drives impact and what distracts you.

Let me state one obvious thing: This order is vital to deliver value. You may be tempted to jump straight to building. I don’t recommend that because you will quickly fall in love with the solution without developing a thorough understanding of the problem.

Slowing down when everyone is speeding up can be your superpower in today’s wild world.

Are You Ready to Become the Product Lead Execs Cannot Ignore?

Have you ever felt that you deserve more recognition than you’ve received?

Many PMs receive a nice tap on the back, but fail to get the promotion they deserve.

I crafted the 100XPM to unlock your growth. This is a serious 3-week mastermind that takes you from PM to product lead. It unlocks your growth.

It’s not a theoretical course like others out there; it’s a mindset reframing that makes you stand out. If you’re serious about your journey, join me and I will help you grow.

You can apply here. And I will review your application.

Here’s what people are saying about it:

I’ve spent the last ten years running very early-stage startups, and as a result, I have had extensive product involvement, but I have never received any formal product training.

100XPM and David’s approach provided me with a wealth of insight into additional tools I could use to maximize my experience and insights, validated what I had already learned along the way, and helped me identify some of the less productive ways I had been thinking about Product Leadership.

Thanks to 100XPM, I feel confident stepping into the Senior Product Leadership role I’ll be starting in the spring.

- Sam Tullman

David’s 100X PM Mastermind was a truly valuable and fun experience.

What stood out most was how thoughtfully the program was designed - not just in the structure of the sessions, but in David’s personal commitment to the growth of each participant.

He brings a great blend of honesty, experience, and genuine care for the craft of product management. David kept the sessions engaging and actionable with a variety of formats: hands-on exercises, real-world examples, and open discussion.

I especially appreciated the small, tight-knit group format, which created a safe and energizing space for reflection, challenge, and growth. There’s always more to learn, but this mastermind gave me tangible tools and perspectives I’ve already started applying.

I’d recommend it to any product manager looking to level up with intention, sharpen their strategic thinking, and surround themselves with a global group of other product managers on the same journey.

- Alex Chiaramonte

Shall we rock together?