A Learning a Day

Notes on LLM RecSys Product – Edition 1

After a few years, I restarted writing a newsletter focused on building technology products. I’m hoping to share a post every month.

---

It’s been years since I last wrote a product newsletter.

My first one – titled Notes on Product Management – was where I tried to make sense of my early years as a product manager. I wrote about roadmaps, storytelling, problem validation – all the messy lessons that come from learning the craft from the many mistakes and helpful counsel along the way.

I thought about writing again when the nature of the work expanded to managing PMs/PM leaders. But I never quite did even if I came close a couple times. Either the work was too consuming, or I just didn’t feel I had something new to say.

But that changed this year when I realized that the job has changed.

Not my title – I’m still a product manager – but the work itself feels entirely different. Especially when I compare how I spend my time now vs. 18 months ago.

Writing long-form has always been my way of thinking. I’ve loved that the word essay comes from the French essayer, meaning “to try.” Essays are how I try to figure things out. This series will be my way of figuring things out.

Also, while I realize that long-form writing feels a bit out of fashion in the era of podcasts and 30-second videos, it still feels like the right medium for reflecting on how the work is changing.

---

Why Write Now – A Counterpoint to AI FOMO

There’s another reason I wanted to write now. Some of the conversation around AI and product management has started to feel increasingly distorted. There’s an entire ecosystem emerging around AI FOMO and AI fear-mongering.

Depending on who you read, AI has already killed the PM role. Or saved it. Or will turn PMs into part-time engineers, designers, hackers, founders – and of course, prompt and eval savants.

Then there are the stories of the AI-native PM – the one who moves at superhuman speed, writing evaluations (“evals, evals, evals…”) before breakfast, shipping beautiful Cursor-powered prototypes and code on GitHub during the day, building incredible vibe-coded side projects in the evening, and operating on a different plane of existence. And, of course, they’re doing all of this at the hot AI-native startup that will save their career while the rest of the PM population goes extinct.

I’d like to offer a counterpoint.

This series will be decidedly anti-AI FOMO – and designed to get fewer clicks. My view is that these takes on the PM craft sound impressive, but are also often, in my opinion, performative.

That’s because they are often disconnected from the real, slow, iterative work of building products people actually rely on.

And because the work itself has changed, I want to focus on that real, iterative practice – not the performative mythology around it.

---

The Fundamental shift – Products Architected Around a Model

There are many takes on how products will be built in the future. I often find myself smiling whenever I see the confidence and certainty behind some of them. Because from where I sit, we’re still at the very beginning of something as fundamental as the Internet – and there’s very little that is certain.

That said, one thing does feel clear: we are moving from products that were architected around deterministic flows – predictable logic and carefully hand-crafted states – to products that are architected around a probabilistic model.

The model is no longer just a feature. Increasingly, the model is becoming the heart of the product. That is a fundamental shift to how we build, ship, and manage products.

---

My Thesis: Why LLM Recsys (Recommender Systems) Are the Core Primitive

At the center of this shift is a single primitive: a model or system that observes context, predicts intent, and recommends what should happen next.

That brings us to my thesis: Most AI-native products share the same underlying loop – and that loop is, fundamentally, a recommender system.

Once we strip away the hype, the interfaces, and the new tools, AI-native products need to: understand context, infer what the user is trying to accomplish, recommend the next step or action, and learn from every interaction.

That is the recommender-system loop. Except that, with LLMs/Large Language models, recommender systems are vastly more flexible and capable. A traditional ranking model used to just choose from a fixed index. An LLM-powered system can reason, generate, act, and orchestrate tools.

What about agents then? An agent is simply the most expressive version of a recommender system: a system that not only recommends what should happen next, but can take that action, explain it, refine it, or even ask questions to clarify intent.

The surface doesn’t matter – it could be a feed, a copilot, a customer service agent, an orchestrator that calls tools, LLM-powered search, or an inbox assistant. They all rely on the same underlying loop.

Even chat interfaces aren’t really “just generation.” Useful ones are making a choice: Should I summarize? Recommend? Rewrite? Plan? Ask a question? Trigger a tool? Take an action? That choice comes from a recommendation policy.

And so, as products shift from deterministic flows to loops that power recommender systems, product teams shift with them. Our job becomes designing the loops – the feedback, guardrails, policies, and signals – that shape how the LLM powered recommender system behaves.

---

A Note on Product vs. Product Management

A quick note on scope – I’m intentionally framing this series around product, not just product management.

Model-centric products blur the traditional boundaries between PM, design, engineering, data science, and AI/ML. When the core of a product is a learning loop – not a static workflow – every function shapes how the product behaves and improves:

• Designers shape the interaction surfaces and conversation design
• Data scientists architect the measurement and diagnosis of improvement opportunities
• Engineers shape the orchestration and safety constraints
• AI/ML teams ensure the model’s capabilities keep getting better
• PMs ensure the policy and evaluation reflect the ideal end-to-end user experience

Everyone touches the loop. Everyone shapes how the product learns. While the disciplines remain deep and distinct, we definitely see a lot of cross-functional flex that, when managed well, result in a better product.

That’s why this series is about LLM Recommender Systems as a product primitive, not a PM technique.

---

What This Series Will Explore

Once you see AI-native products as learning loops rather than deterministic workflows, a number of implications follow:

• Agents aren’t a new paradigm – they’re Recsys that can act
• Product policy becomes as important as the PRD – a key step to designing a great human-in-the-loop system
• Teams evolve around capabilities (“atoms”) and surfaces (“molecules”)
• Evaluations shift from one-time tests to continuous diagnostics
• Velocity depends on how well the system learns, not how fast teams ship

These are the ideas I’ll explore in upcoming editions – one concept at a time, likely once or twice a month.

This isn’t a guide or a manifesto from someone who’s figured it out. It’s a set of reflections written in motion – from someone who’s building, operating, and trying to understand this shift in real time.

There’s a lot changing, and a lot we don’t yet know. My goal is simply to ground in first principles, learn publicly, and share the process along the way.

If any of this resonates, I’d love to hear from you. The best part of writing is the dialogue that follows.

Building a large product is often like putting together a massive jigsaw puzzle.

You have people working on different ends – some on the left corner, some on the right, some deep in the middle. You begin with the end in mind, imagining what the final picture might look like, and then you start connecting the pieces.

In the middle of the journey, it’s messy. You’re just working through bottlenecks – sometimes on one edge, sometimes in the center, moving wherever the next piece demands your attention.

And then, every once in a while, there’s that magical moment when you zoom out. Suddenly, you can see the picture taking shape. The edges fit, the colors align, and the hours of agonizing over small details start to make sense.

It’s in that moment – when all the effort, doubt, and iteration come together – that you feel a quiet rush of joy.

Goosebumps, even.

It doesn’t last long. You admire the finished puzzle for a few seconds before you’re on to the next one. But if you’ve done it right – if you’ve enjoyed the process, the collaboration, the craftsmanship – those few moments make it all worth it.

One of the more important work lessons is to be someone who knows how to say no.

Every once in a while, you will see a horrible idea being proposed. When that happens, you need to be unequivocally clear that your answer is a “no way.”

In all the other cases, the goal isn’t to shut an idea down with a “no”. It’s to help others understand what it would take to get to yes, and to help them get there.

It’s an easily forgotten truth: the job isn’t about pointing out why something won’t work. It’s about using critical thinking and healthy skepticism to strengthen the eventual solution.

Saying no is easy.

Helping someone get to a better yes – that’s leadership.

It is amazing how the answer to so many conundrums in our lives and careers becomes so much clearer when we are able to answer the question – what are you solving for?

One of the signs of being well-rested is that the body doesn’t need an alarm to wake up.

It knows enough to wake us up early when there’s work that needs to be done.

Rest well enough, and our body and its internal clock work in sync.

Beware spending your life seeking validation from people who don’t really care about you.

It’s one of the easiest traps to fall into – shaping your choices and maybe even your dreams around the approval of people whose opinions won’t matter in the long run.

When in doubt, focus on the validation of the person you see in the glass.

“You may fool the whole world down the pathway of years,
and get pats on the back as you pass.
But your final reward will be heartaches and tears
If you have cheated the man in the glass.” – The Man In The Glass by Peter Dale Winbrow Sr

Someone said recently that their partner has a simple maxim – in the final analysis after this life is done, the only question you’ll be asked is – did you make the most of your gifts?

It is a beautiful question.

A reminder to do the best we can with what we have where we are.

There can never be enough such reminders.

I came across this thought provoking post titled “Why Architecture Matters.”

It starts by explaining the “why” behind Gothic architecture.

Staying in London, take the example of Westminster’s great Gothic church. Its pointed arches and lofty spires give the sense of upward movement. Its wide base adds a feeling of groundedness and solidity. Its fine embellishments like stained glass and carved arches suggest that even on the grandest of scales, no detail is too small to be overlooked.

This style of building emerged in Europe in the 12th and 13th centuries, and is a perfect reflection of the deepest beliefs of Europe in the late middle ages. Medieval architects took for granted that man’s purpose was to journey toward heaven, which is why they built a sense of upward motion into their cathedrals. Yet they also knew that in order to do so, you must stay grounded in your earthly life — and thus they gave their buildings a solid foundation, both functionally and visibly.

Most importantly, however, they believed that beauty has moral power. The designers wanted to create a building that would ennoble and inspire every person who walked in. They filled their churches with painstaking detail so that every aspect offered an encounter with the kind of beauty that draws man toward the divine.

It then makes a powerful point – Whether a building’s designers are conscious of it or not, architecture always tells the story of a culture’s values. That’s why if you want to know what your culture believes today, you should look at what it builds.

In doing so, it examines the difference between architecture in the US vs. Hungary. American cities are designed around the car. Perhaps it means efficiency is valued more than beauty – a result of a mindset focused on productivity.

Prague, on the other hand, looks and feels different – a result of a culture that prioritized beauty, prayer, connection to the past, and staying connected to one’s community.

It ends with a beautiful note.

In an age that claims all beauty is in the eye of the beholder, it’s worth reminding ourselves that architecture is never neutral. In fact, it is arguably the best physical embodiment of what a culture believes and how it lives. It is shaped first by our values, and then reinforces those same values in us.

Next time you go out, be sure to take a close look at your house, your church, your pub, your city. Try to read the values that underlie the physical building. What do you see?

And perhaps more importantly, what would you like to see instead?

It resonated.

I find it useful to contemplate your own insignificance in the universe from time to time.

Maybe one person from any given generation will be remembered a thousand years from now – that is, if humans are lucky to still inhabit the planet a thousand years from now.

For the rest of us, our time here is brief. We might influence a few members of the next generation, perhaps even the one after that, if we’re lucky.

But not much more.

When you sit with that insignificance, it puts everything into perspective.

Nobody outside really cares. The striving, the titles, the noise – all of it fades quickly.

What’s left is you, and your ability to find balance – between striving and equanimity, between contribution and peace.

Because in the final analysis, what counts is simple:

that you did your best,

with what you had,

and made the most of the time you were given.

The more I experience life, the more I realize that happiness is a combination of two things – a sense of contribution and a sense of peace.

They can feel like opposing forces.

Contribution pulls us outward – toward action, effort, and usefulness.

Peace pulls us inward – toward acceptance, stillness, and letting go.

But they’re not opposites. They’re complementary.

We’re wired to contribute – to feel useful, to make a difference in some way. Yet peace comes from releasing attachment to outcomes, from not wishing to be someone else, somewhere else, or chasing validation.

Happiness, then, is the balance between the two: where we give ourselves to a cause that’s bigger than us and where we possess the perspective to let the universe unfold as it should.