Netflix CPTO on AI and the future of product and tech roles | Elizabeth Stone

Everyone can do everything now, and it’s creating real whiplash on teams—PMs coding, designers drafting specs, engineers shaping product—so what exactly is my job anymore?

We’re in the messy storm before the shape emerges, and the answer isn’t to put AI back in the box; experiment, but pair it with clear guardrails, trusted data, and the reminder that humans stay accountable for outcomes.

Quick rewind for context: today my guest is Elizabeth Stone, Netflix’s product and technology officer; her last visit became one of this show’s most-loved episodes, and now we get to trace how AI changed the game.

PM, design, and data science can now push ideas to testable prototypes faster, using AI to mine years of research and experiments, while engineering still owns the how—scale, reliability, and quality.

If everyone’s a builder, do we still need distinct functions, or should we collapse into one generalist track?

Craft matters more than ever; great engineering, data science, and creativity are scarce, and while tools broaden our reach, deep discipline excellence is still the differentiator.

Are you shifting hiring toward certain roles as AI rises?

We’re leaning into systems thinkers and platform builders—shared infrastructure, paved paths, and source-of-truth data—so teams and agents can move fast without shipping a Frankenstein product.

Is the push for platforms mostly about speed, or something else?

It’s leverage and safety: encode identity, security, quality, and data access in the platform so thousands of people and many agents can build without hunting tribal knowledge.

Some argue the design process is dead in an agent-driven world; do you agree?

We should enable more people to build well, but for our highest-impact bets, deep design remains essential to make complexity disappear and deliver a seamless, distinctly Netflix experience.

What skills are fading, and which are rising?

Narrow specialization is less useful than adaptable talent that spans domains, though some rare experts remain vital; the mindset is learn fast, zoom out, and evolve your toolkit.

How can people build real systems thinking instead of just saying the words?

Zoom out one level on every problem to test assumptions, think like your manager about how pieces fit together, and leave systems stronger for colleagues who build next.

Have you changed career ladders for AI, or handled it another way?

We overlaid a company-wide expectation of AI fluency—curiosity, judgment on when to use it, and hands-on skill—reflected in hiring and interviews where candidates can use AI tools.

Beyond coding, where is AI moving the needle at Netflix that people might miss?

Data distillation supercharges analysis and decision-making, and across content we use AI for previsualization, post-production, localization, and at-scale creative assets, including tech from our Interpositive acquisition.

You were early to ML, and personalization only got harder as the catalog grew; how does that shape today’s approach?

Personalization is core to matching the right title to the right member across film, TV, games, live, and more, and our long ML history gives us a head start as the tech leaps forward.

Your culture sounds like the AI labs—high agency, talent density, experiments, and top-of-market pay.

We think of it as excellence as the operating system—push decisions deep, hire for judgment, embrace risk, and keep process light so great people can do their best work.

What are the pillars leaders need to make that real, not just a poster on the wall?

Start with talent density, prize outcomes over ego, let people make consequential calls and learn, run blameless retros, and resist layering process when things get tough; be highly aligned and loosely coupled.

Is the Keeper Test still central, and how has it evolved?

It sparks frank, frequent feedback—often celebrating impact and sharpening strengths—and when the answer is no, we either outline a path to get there or make a hard call sooner.

With fierce competition, how do you win top talent today?

We attract builders who love applying tech at global scale to entertainment; if you want to shape what people use every day, that mix of product, tech, and storytelling is uniquely energizing here.

What about juniors in an AI-first world—how do they grow?

We hire interns and new grads, invest in mentorship and craft quality, and value their native fluency with new tools and shifting consumer behavior, while expecting them to own the outputs they ship.

Looking ahead, will engineers still need to code, or just understand systems?

Even if agents write more code, humans must understand how systems work to judge quality, debug, and improve; today’s agent code can feel opaque, so we’re developing new practices to build confidence.

How will entertainment itself change over the next few years?

Entertainment will span more formats—live, games, podcasts, mobile, and TV—and the job is to make the journey personal, immersive, and easy, so you can flow from a podcast to a series to a game without friction.

Hollywood has mixed feelings about AI; how do you work with creators across that spectrum?

We enable creators however they want to work, from traditional methods to new AI-enhanced workflows, and we partner flexibly because the future will include both familiar and entirely new formats.

Will fully AI-made shows win hearts, or do humans stay central?

Human storytelling is the backbone—AI will amplify and accelerate production, but people carry the emotion that makes stories resonate.

Before the lightning round, any closing thoughts to underline?

This is a thrilling time to build in entertainment, and while the tech is amazing, the mission is simple—create products and stories people love.

Lightning round: favorite books, a recent show, a product you love, and a motto you live by.

Into Thin Air and Liar’s Poker are go-tos, Remarkably Bright Creatures moved me, Eight Sleep helps me function, and I live by noticing something good every day and giving the last five percent.

You’re cycling the last week of the Tour de France route—racing or cruising?

It’s not a race, but those mountain stages are no joke, so I’m training to enjoy it and not be last.

Where can people find your work, and how can listeners help?

Check the Netflix Tech Blog, try our new live events, games, and the Clips feed on mobile, and send feedback so we can make it better.

Thanks for coming back on; if you enjoyed this, subscribe, rate the show, and find past episodes at Lennyspodcast dot com.