Oct 10, 2025 · 1h 21m · lennys-podcast
Inside Google's AI turnaround: AI Mode, AI Overviews, and vision for AI-powered search | Robby Stein
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this in-depth interview, Google VP of Product Robby Stein joins host Lenny Rachitsky to break down Google's rapid AI turnaround, the development of AI Mode and multimodal search, and the timeless product management principles behind scaling consumer hits at Instagram and beyond.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Lenny holds 26.1% of the talking time here. How this is scored →
Lenny as informed peer 4.7 Guest teaching 4.1 Guest disagreement 1.2 Lenny pushing back 1.5
speaking balance: gold is Lenny, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 34:19 Robby reframing format innovation vs copying
Robby firmly dismisses the accusation that Instagram simply stole Snapchat's product, arguing that core formats naturally become universal primitives across products just like feeds.
Hardest push from Lenny ▶ 34:05 Lenny confronting Robby on Instagram copying SnapchatLenny directly challenges Robby on the industry backlash and accusations from founders that Instagram bluntly stole Snapchat's core innovation.
Biggest teaching moment ▶ 8:24 Robby reframing the death of search premiseRobby dismantles the narrative that AI chatbots are killing search, educating the host on the vastness of consumer queries and demonstrating that AI is expansionary rather than zero-sum.
Lenny holds their own ▶ 20:58 Lenny connecting conversational AI to product delight frameworksLenny actively demonstrates deep PM mastery by linking Robby's insights on natural language models directly to past product leadership lessons from Google Meet.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Lenny as informed peer | Guest teaching | Guest disagreement | Lenny pushing back | Why |
|---|---|---|---|---|---|---|
| Google's Shift in Urgency and Gemini's Momentum | 5 | 4 | 1 | 2 | Lenny opens by pointing out Gemini's rapid climb to the top of the App Store and questions what structural shift took place at Google. Robby explains the compounding internal focus, urgency, and cross-team collaboration with DeepMind. | |
| The Expansionary Nature of AI in Google Search | 4 | 6 | 2 | 3 | Lenny brings up the prevailing narrative that AI chatbots mean the death of traditional search. Robby educates Lenny by showing that search query diversity is massive and that AI acts as an expansionary force rather than a replacement. | |
| Shifting from Keyword Queries to Natural Language | 6 | 5 | 1 | 2 | Lenny invokes Alex Rampell's startup distribution versus incumbent innovation quote. Robby explains the mechanics of query fan out and why original informational value remains decisive. | |
| Developing Human-Like Interfaces in AI Products | 7 | 3 | 1 | 1 | Robby describes the transition toward natural language steering without fine-tuning. Lenny demonstrates his expertise by connecting this concept to Nasreen Shingel's philosophy on human-centric software design. | |
| Embodying Relentless Improvement and Productive Dissatisfaction | 4 | 4 | 1 | 1 | Robby reflects on productive dissatisfaction, sharing personal anecdotes including his wife's feedback and Tony Fadell's fruit sticker talk. Lenny facilitates the narrative smoothly. | |
| Identifying User Pain Points to Create AI Mode | 5 | 5 | 1 | 2 | Lenny asks how to balance qualitative craftsmanship with rigid metric goals. Robby details how observing users appending the letters AI to queries inspired AI Mode and how retention curves guide interventions. | |
| Lessons from Building and Launching Instagram Stories | 6 | 5 | 4 | 6 | Lenny directly confronts Robby on Instagram copying Snapchat Stories and the ensuing backlash from founders. Robby defends the decision, arguing formats become universal primitives like news feeds and that serving user needs supersedes format purism. | |
| Frameworks for Unlocking Growth in Mature Products | 5 | 5 | 1 | 1 | Robby details his playbook for mature products using Jobs to be Done and making new paradigms complementary rather than destructive. Lenny prompts for actionable frameworks. | |
| Managing S-Curves and Resource Allocation for Big Bets | 6 | 3 | 1 | 1 | Lenny crisply synthesizes Robby's resource allocation strategy into managing overlapping S-curves. Robby confirms the quantitative framework of tracking diminishing marginal returns. | |
| Sponsor Message: Orkes Conductor Workflow Orchestration | 4 | 4 | 1 | 1 | Following the sponsor read, Robby walks through the early zero-to-one prototyping of AI Mode with a small cross-functional pod and trusted testers. | |
| Google's AI Speed and Positioning AI Mode | 5 | 4 | 1 | 2 | Lenny highlights Google's speed and asks how AI Mode differentiates itself from ChatGPT and Perplexity. Robby articulates that AI Mode focuses strictly on authoritative informational utility rather than companionship or open-ended creative tasks. | |
| Three Core Product Philosophies for Product Builders | 5 | 5 | 1 | 1 | Robby outlines his three core principles: deeply understand people, analytical rigor on root causes, and clarity over cleverness. Lenny engages constructively and validates the naming clarity of AI Mode. | |
| Transforming Instagram Close Friends from Flop to Hit | 5 | 6 | 1 | 1 | Robby gives a comprehensive breakdown of turning Close Friends from a low-engagement flop into a core feature by analyzing list sizes and emotional Jobs to be Done. Lenny adds context regarding the finsta cultural trend. | |
| Reassessing Lean Teams and Scaling Hard Technical Bets | 4 | 5 | 3 | 2 | Robby takes a contrarian stance against tech Twitter dogma regarding ultra-lean teams, arguing that hard technical bets and enterprise-scale consumer problems require substantial resource backing once conviction is reached. | |
| AI Corner: Visual Search and Image Inspiration | 4 | 4 | 2 | 1 | In AI Corner, Robby discusses visual multi-turn search for aesthetic inspiration. Lenny playfully compares it to Nano Banana and Pinterest, while Robby clarifies the exact architectural difference. | |
| Fostering Curiosity and Raising AI-Native Kids | 6 | 3 | 0 | 1 | The conversation shifts to parenting and curiosity. Robby talks about voice search live with his kids, and Lenny contributes a practical DIY example of an AI-powered voice pirate toy. | |
| Lightning Round: Favorite Books, Entertainment, and Pillows | 4 | 2 | 0 | 0 | A relaxed lightning round discussing fiction recommendations like Project Hail Mary, TV shows like The Bear, and Robby's enthusiasm for the Purple Pillow. | |
| Life Motto and Pitching Justin Bieber for Stamped | 3 | 4 | 0 | 1 | Robby recounts pitching Scooter Braun and onboarding Justin Bieber for his startup Stamped by flying to LA on zero notice to illustrate bias for action. | |
| Connecting with Robby Stein and Providing Feedback | 2 | 1 | 0 | 0 | Standard wrap-up where Robby invites direct user feedback on Google search features and Lenny closes the episode. |
Statements from this episode (30)
Rachitsky: Google Gemini reached number one on the App Store above ChatGPT
“So just last week, Gemini, Google Gemini hit the number one app in the App Store. I have it right here. It's still number one in the App Store. It's above ChatGPT.”
Stein: Product momentum compounds from daily improvements rather than single changes
“I think that a lot of times people ascribe, you know, a lot of momentum to a one time change or a single person, right? I find a lot of this is actually this compounding effect. And you think about just every month, Like ruthlessly improving the product or the…”
Stein: AI expands overall search queries without hurting core Google Search
“And what we're finding is that AI is expansionary. And so there's actually just more and more questions being asked and curiosity that can be fulfilled now with AI. And so that's where you get the growth. And so all like the core Google search isn't really cha…”
Stein: Google Lens visual searches grow year-over-year across billions of queries
“Google lens, one of the fastest growing products out there, it's growing. Year over year increase in visual searches, which is already like a massive scale. It's like billions and billions and billions of searching in that way.”
Stein: Google Shopping has 50B products updated 2B times hourly
“There's fifty billion products in the Google shopping graph, for instance, they're updated two billion times an hour by merchants with live prices, right? You have two and fifty million places in maps.”
Stein: Users are bringing complex calculations and visual questions to Google Search
“What we found is that people Are asking these questions in Google, like they're trying to get this out of Google. And so if you can just have an AI that's powerful enough to answer a really hard calculation, someone's trying to figure out or like take a pictur…”
Stein: The biggest shift in Search is users asking long, complex questions
“That's the biggest shift we're seeing people asking real long, hard, complex questions.”
Stein: Google AI executes query fan-out with dozens of background searches
“When our AI constructs a response it's actually trying to, it does something called query fan out. Where the model uses Google search as a tool to find, to do other querying. So maybe you're asking about specific shoes. It'll add and append all of these other …”
Stein: Core SEO quality signals remain extremely valid for AI search
“There's these best practices that I think still do largely apply because it's gonna ultimately come down to an AI is doing research and finding information. And a lot of the core signals, is this a good piece of information for the question? They're still vali…”
Stein: AI models increasingly do not require heavy fine-tuning for sophisticated outcomes
“I think it's gonna open up a lot of this democratization of accessing these models and building incredible things. Cause you don't even need to do a lot to get the most sophisticated outcomes. Increasingly. I don't think you need to do a lot of this heavy duty…”
Stein: Users typed 'AI' at end of Google queries to trigger answers
“We actually saw people in the query stream. Putting the words AI at the end of their queries because they're trying to like get the AI to like do the thing.”
Stein: Contorting an established core product to do something new fails
“Whenever you try to mix a core product, that's very cemented in someone's mind and physically looks a specific way. And you're trying to make contort it to do something new. It's usually a bad recipe.”
Stein: Snapchat initially did not permit photo uploads from camera roll
“Snapchat at the time didn't allow you to upload photos. It was like, you have to use a snap camera.”
Stein: Successful product formats quickly become primitives every app must adopt
“Facebook probably created the modern feed, but there's a feed for every single product, right? And there's a LinkedIn feed and there's a feed for DoorDash. You know, it's not like these things become core primitives quickly and formats. And then at the end of …”
Stein: Instagram sharing shifted from public feeds to Stories and DMs
“You know, certainly in Instagram, there's been a big shift over the years of sharing into public, very large broadcast posts and feed into these more lightweight formats, like stories and DM, actually private sharing as well.”
Stein: New product formats expand platforms rather than replacing core features
“In many ways, the AI version of Google and stories and reels, they're all kind of similar in that they're new formats in the world that people are expecting and wanting more of. And by adding them, It becomes complimentary, not replacement. And in both cases, …”
Robby Stein: Diminishing returns make adding 50 people useless on mature products
“And you get to these points of just diminishing marginal return in every system where it feels like you could put 50 people on this project. Like it's just not gonna dramatically move the needle. And so part of it is this bottoms up thing with your own team be…”
Stein: New product growth engines yield 10% to 20% wins per optimization
“And then when you land a new thing that creates this new little growth engine And then you put people on it and you optimize it because you get, you're getting big, like each change is like 10% win, 20% win, four percent win. And it's clearly like still has so…”
Stein: Conversational AI search cannot fit within core search UI
“It's gonna be really hard to build all of that within the construct of the core search experience.”
Stein: Google originally built AI Mode with 5 to 10 people
“I mean, this was probably like five to 10 people worth of people originally.”
Stein: Google tested AI Mode with ~500 people before Labs
“And then we did bring it before labs actually to trusted tester group. There were maybe like, 500 people externally that we added onto it, and we had pings with them.”
Stein: Next year of AI products will set user habits for years
“I think people believe, and I certainly believe that the next year or so of product is going to kind of establish how people use the next wave of products from, for many years.”
Stein: Google AI Mode focuses on information, not productivity or creativity
“We are less focused on, you know, creativity, productivity, like upload a spreadsheet and like output graphs for me. Like we're not focused on that. Like we're really focused on what people use Google for and making an AI for that so that you can come to Googl…”
Stein: Instagram Close Friends initially failed due to 'best friend' mistranslations
“There was this, there's a story in, in close friends at Instagram where It just totally failed at first on, on in, in a bunch of, just when we shipped it. And it turned out that we looked at the data and people were only adding one close friend to their list b…”
Stein: Instagram Close Friends succeeded once users added 20 to 30 people
“It was working really well for people who added 20 to 30 people to their list. Cause what would happen is you put 30 people on your list and then two of them would write back to you on DM. And now you have closed the loop and you feel connected to those people…”
Stein: The cult of lean causes teams to under-invest in technical breakthroughs
“There's kind of this cult of lean, scrappy, fast, throw away your product quickly, keep moving. And I think at some level it's true for internal conviction, but to build a product that works for a lot of people that is based on a techno technological breakthro…”
Stein: Friends won't use a product past 30 days out of loyalty
“I've, I found out very quickly building startups that if you put 20 friends on something, they're not gonna do you that many favors. Like they're not gonna use a product every single day. Cause they're your friend, like 30 days in, 60 days in, 90 days in. They…”
Stein: Google is rolling out multi-turn visual search in AI Mode
“There's a visual version of AI mode that basically we talked about for at IO. And so you can reference some of those keynotes, but that that's in the process of being rolled out. And so you're going to be able to now ask like, you know, what's a mid century, m…”
Rachitsky: Pinterest is in trouble as AI visual search emerges
“Pinterest is in trouble. Feels like this is what people use Pinterest for. Here's all the inspiration. Now it's just AI doing it all.”
Stein: Google launched Search Live conversational voice out of Labs
“Funny enough, we actually just launched search live actually out of labs this week. And so you can talk to search in a live AI setting, which is conversational voice, voice on when you're driving, you can just talk all the knowledge I talked about, what you ca…”