Lenny's PodcastHow to ship hardware in the AI era | Caitlin Kalinowski (Apple, Meta, OpenAI)
At a glance
WHAT IT’S REALLY ABOUT
Shipping AI-era hardware: robotics, supply chains, safety, and teams at scale
- VR’s biggest impact may be as a technology stepping stone—advancing spatial sensing, SLAM, and depth tech that now directly powers robotics, autonomy, and “physical AI,” even if VR remains niche.
- AR glasses are still a credible long-term interface, but mass-market products are blocked by manufacturing readiness (e.g., waveguide and micro-LED yield/cost) and by unsolved input/interaction challenges in public settings.
- Robotics enthusiasm is rising because AI work “behind the keyboard” may eventually saturate, pushing the next frontier into the physical world—manufacturing, autonomy, drones, and robots operating in real environments.
- Scaling robots is less about flashy demos and more about reliability, safety around humans, and fragile supply chains—especially magnets, actuators (motors), silicon, and memory—where a single shortage can force catastrophic redesigns.
- High-performing hardware organizations win by setting stable early KPIs, designing the riskiest constraints first, iterating most on the parts customers touch, and building teams that combine seasoned builders with AI-native young talent.
IDEAS WORTH REMEMBERING
5 ideasTreat hardware like you only get a few “compiles,” ever.
Unlike software, hardware iterations are slow and limited; each major build can consume months, and after mass production you can’t “OTA” your physical mistakes, so rigor and conservatism must be baked into the process.
Lock goals (KPIs) early; changing them midstream is brutally expensive.
If you shift targets like price, weight, or performance late, you often trigger redesign cascades across mechanics, electronics, manufacturing, and reliability testing—burning schedule and competitive timing.
Design the riskiest pinch points first, not the parts you already know.
Great hardware architects start where failure is most likely (fit, hinge cable routing, tolerance stacks, thermal limits) to de-risk feasibility before polishing the obvious components.
Iterate disproportionately on what the customer touches most.
Trackpads, keyboards, grips, or human-contact surfaces require more cycles because feel and reliability dominate perceived quality; less-touched internals can often tolerate fewer iterations.
Humanoid robots are still advanced prototypes; safety is the gating constraint.
Strong robots near humans require proven safety via reduced mass, softer compliance, and intent signaling; many existing systems effectively require humans to keep distance, which blocks consumer/home adoption.
WORDS WORTH SAVING
5 quotesThere's a dawning realization, especially in the lab, the acceleration is going so vertical that what you can do behind a keyboard with AI is gonna saturate. When that happens, the next frontier is the physical world, robotics, manufacturing, industrialization.
— Caitlin Kalinowski
In hardware, we only get to compile our code, quote-unquote, like four or five times.
— Caitlin Kalinowski
I think there's probably more change in war than there is in consumer electronics in the next two years.
— Caitlin Kalinowski
If you walk into a room and a robot's just like... Like, it's creepy. You want these devices to be non-threatening, appear soft, reactive to you.
— Caitlin Kalinowski
Sam is really good at saying, "Why not more? Why not 100x or 10,000x? You're thinking too small." For Steve, the bar he held for the company, for technical talent, and for excellence was not wavering.
— Caitlin Kalinowski
High quality AI-generated summary created from speaker-labeled transcript.
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome