Saturday, July 25, 2026

Hi {{first_name | there}},

I have to tell you: I am feeling a little validated.

System Addict is two issues old. Then, this week, Lenny's Podcast released an interview titled Why Netflix is betting on systems thinkers, not specialists, in the AI era.

I’m not suggesting Lenny Rachitsky or Netflix built their editorial calendar around my two-week-old newsletter. That would be ambitious, even for me. But I’m happy to enjoy the timing.

If you’ve been reading from the beginning, take the small win with me. We got to this conversation early.

Starting next week, I’m going to make System Addict more hands-on. I will begin taking apart some of the systems I use to run my business, including what they do, how the pieces connect, and where they (sadly) still break.

You will also receive a separate subscriber email later today with a bonus report: Nine Operating Systems: What 2,907 LinkedIn posts reveal about the systems behind nine top creators. I analyzed the public content systems of nine business creators, the different publishing machines visible in their work, and the patterns worth testing in your own personal-brand system. It’ll be a fun read, so please share it with anyone who might enjoy it.

For now, I want to stay with the Netflix conversation because Elizabeth Stone, Netflix's Chief Product and Technology Officer, gave one of the clearest explanations I have heard for why systems thinking matters more as AI gets better.

Stone describes the current moment as the messy phase that follows any major technology shift. This one has led to job descriptions spilling into other roles.

Product managers can prototype. Designers can draft product requirements. Engineers can take on more product thinking. Agents give each of them more reach beyond their core discipline.

That flexibility is useful. Teams can test an idea before the usual handoff becomes a bottleneck. People can search years of research, experiments, and company knowledge without tracking down the one colleague who remembers where everything lives.

It also produces a lot more output.

A company can suddenly have far more prototypes, analyses, and code changes moving through the organization. More people can build, but they do not automatically share the same definitions, data, security rules, quality bar, or understanding of the customer problem.

AI removes many local bottlenecks while leaving coordination unsolved.

Stone says it plainly: "We need more systems thinkers in a world with AI."

When more people and agents can build, organizations need common infrastructure that gets each of them most of the way to a safe, useful result. They need source-of-truth data. They need reusable components, design systems, access rules, tests, and clear quality standards.

Otherwise every new builder has to rediscover the same institutional knowledge:

Engineering needs people who can look across business domains and identify the shared building blocks underneath them. Design needs templates and systems that help non-designers create something coherent without erasing deep design work on the highest-stakes problems. Data science needs common definitions and trusted data so faster analysis doesn't become faster confusion.

This is systems thinking as a practical response to scale. It means seeing how local decisions interact, identifying what should be shared, and building foundations that keep the work pointed toward the same outcome. Higher velocity makes that ability more valuable.

Stone doesn't make salary predictions, and one Netflix interview is of course not a labor-market study. Still, I'm willing to make the bet that more of the professional premium will move upstream.

AI can help produce the first draft, prototype, analysis, or block of code. As that execution gets faster, more value sits in the decisions around the output:

These are systems-thinking questions. Each one pushes beyond the immediate artifact and into the relationships around it: the outcome, repeated use, dependencies, scale, and accountability. They reward people who can connect a local task to the larger system and still make something useful.

Deep craft still matters. Stone is explicit that excellent engineering, data science, and creativity remain scarce. She separately argues that deep design work still matters on the highest-stakes consumer problems. Netflix still needs specialists in areas where only a small number of people understand the technology or domain.

Narrow expertise becomes fragile when it can't zoom out and look around. Craft plus range is more durable: someone who can go deep, connect the work to the larger system, and learn when the old method stops fitting the problem.

AI gives more people reach. Systems thinking decides whether that reach compounds or creates an expensive mess. That is the premium I'm betting on: staying close enough to do great work while zooming out far enough to make that work dependable, scalable, and resilient.

Stone's advice for building that range is simple. Before solving the task in front of you, move one level up and inspect the broader space around it. Don't disappear into a three-week strategy exercise. Take one beat, ask the wider questions, then return to the work.

I translated that move into four questions you can use this week. For your next meaningful task, write down:

That short exercise can help add shape to the solution, and your counterparts will likely thank you for taking a systems approach because it can help unearth hidden considerations.

A request for a dashboard may become a question about which decisions the team cannot make today. A request for an AI agent may reveal that source data is inconsistent or unreliable. A request to automate one step may expose a broken handoff on either side of it.

Systems thinking shouldn't get in the way of getting the task finished. But it will make you less likely to solve it in a way that creates the next problem.

The fourth question above, the quality owner, connects directly to another point Stone returns to throughout the interview: the human remains responsible for what the system produces.

An agent can write the code. It can do the analysis. It can help someone work outside their original specialty. The person and the organization still own the quality, the interpretation, the security, and the consequences.

This is why Netflix is treating AI fluency as an expectation across roles and seniority levels. Fluency is not measured by how often someone opens an AI tool. It includes knowing when the tool is useful, when its output needs review, and enough about the underlying system to recognize when something has gone wrong.

The same logic applies to early-career talent. Faster production does not remove the need for mentorship, testing, review, and learning what good work looks like.

Accountability only works when people have room to exercise judgment. Stone describes Netflix's culture as "excellence as an operating system." The goal is to give talented people context, agency, and accountability, then let them make decisions and learn quickly.

Systems thinking shouldn't bog down every process with a checklist. Netflix has found that more process can consume more time without improving the outcome. Useful systems encode what must be shared, including the goal, source of truth, guardrails, feedback loop, and owner, then leave room for judgment.

The aim is enough shared structure to preserve quality without turning every decision into an approval queue.

This week, try the one-click-up practice on one task. If you have questions, or want to share how you're seeing systems thinking play out in your organization, send me a note.

See you next issue.

Jamaur