Sep 16, 2025 · 40m · y-combinator
Aaron Levie: Why Startups Win In The AI Era · Y Combinator
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 Y Combinator AI Startup School talk, Box Co-founder and CEO Aaron Levie joins partner David Lieb to discuss why early-stage AI startups possess a unique advantage over legacy incumbents during the current generative AI wave. Levie shares insights on enterprise AI opportunities, evolving SaaS business models, agent workforce economics, and practical strategic advice for young founders.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
The partners as informed peer 3.2 Guest teaching 5.3 Guest disagreement 1.7 The partners pushing back 1.8
speaking balance: gold is the partners, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 28:40 Dismissal of in-house enterprise software builds
Aaron Levie strongly rejects the premise that enterprise companies will use AI to replace vendors, mocking recent public examples like Klarna as useless PR and fireside fodder.
Hardest push from the partners ▶ 12:39 Challenging AI job growth with Amazon layoffsDavid Lieb counters Aaron Levie's optimistic claims about AI augmenting human labor by directly raising Amazon's breaking announcement of corporate headcount reductions.
Biggest teaching moment ▶ 28:10 Core vs. Context enterprise frameworkAaron Levie breaks down Geoffrey Moore's Core vs Context doctrine to explain why companies will never take on the legal and technical liabilities of building non-core systems internally.
The partners hold their own ▶ 24:14 Host draws on Google Photos economicsDavid Lieb demonstrates his domain expertise by citing his leadership at Google Photos, illustrating how proprietary software commands 90%+ gross margins despite commoditized underlying storage.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The partners as informed peer | Guest teaching | Guest disagreement | The partners pushing back | Why |
|---|---|---|---|---|---|---|
| Opening Highlights on the AI Startup Opportunity | 4 | 5 | 2 | 3 | David Lieb prompts Aaron Levie on the parallels between the early cloud transition and current AI adoption. Lieb pushes back moderately by questioning why enterprise leaders are convinced of AI despite a lack of proven deployed solutions, leading Levie to detail the cultural awareness shift driven by ChatGPT. | |
| Unlocking Enterprise Value in Unstructured Data with AI Agents | 3 | 6 | 1 | 1 | Lieb asks how AI expands Box beyond cloud storage. Levie delivers an educational breakdown distinguishing structured database records from unstructured documents and explaining how agents unlock enterprise knowledge. | |
| AI Agents, Labor Economics, and Job Augmentation | 3 | 5 | 2 | 1 | Lieb introduces the concern of AI labor displacement. Levie rejects conventional media alarmism, framing AI agents as tools that eliminate non-strategic overhead and unlock economically infeasible backlog work. | |
| Why Startups Gain Greater Leverage Than Corporate Incumbents | 5 | 5 | 3 | 6 | Lieb pushes back directly against Levie's optimistic job thesis by citing Amazon's announcement of upcoming AI-driven headcount cuts. Levie concedes the enterprise efficiency motive but pivots to argue that startups will use AI leverage to accelerate hiring and expansion. | |
| Identifying New Nouns and Verbs for B2B AI Startups | 3 | 6 | 1 | 1 | Lieb asks how startups can compete against established incumbents like Box. Levie presents a historical analysis of solved consumer and enterprise verbs between 2008 and 2014 versus the greenfield opportunities emerging today. | |
| Transitioning from Per-Seat Licensing to Consumption Pricing | 4 | 6 | 1 | 2 | Lieb inquires about business model evolution away from per-seat SaaS licensing. Levie explains outcome-based consumption pricing while cautioning about the necessity of maintaining subscription predictability. | |
| Deflationary Supply-Side Economics and Value Capture in AI | 6 | 5 | 2 | 4 | Lieb questions whether customers will tolerate significant markups over raw token costs and later challenges margins against infinite competition. Lieb cites his own experience building Google Photos, validating high margins on commoditized underlying storage. | |
| Core vs. Context: Why Enterprises Will Not Build In-House Software | 3 | 7 | 3 | 2 | Lieb raises an audience concern that enterprise clients will build internal custom software using AI rather than buying SaaS. Levie educates using Geoffrey Moore's Core vs Context framework and dismisses bespoke internal tooling hype like Klarna's as impractical novelty. | |
| Strategic Mindsets and Four Pillars for Early-Stage Founders | 2 | 5 | 1 | 0 | Lieb asks Levie for overarching tactical advice for early-stage student founders. Levie emphasizes canonical business literature, co-founder dynamics, market tailwinds, and capitalizing on the current AI wave. | |
| Audience Q&A: AI Innovations in Data Storage and Lifecycle Management | 1 | 4 | 2 | 0 | In an audience Q&A portion, audience members ask about AI in raw data storage and the meaning of life. Levie explains that raw storage is solved while lifecycle management and higher-level data intelligence represent true AI innovation. | |
| Audience Q&A: Craft and Design Quality in Enterprise Software | 1 | 4 | 1 | 0 | An audience member asks about the role of design and craft in enterprise software. Levie notes that enterprise software buyers historically tolerated utilitarian design, but encourages founders to hold a high aesthetic standard. |
Statements from this episode (17)
Levie: Slow Enterprise AI Adoption Is Due to Change Velocity, Not Disbelief
“So conversely with AI you're no longer really having to convince people that AI is the future. Everybody tends to be bought in, in the enterprise segment. There's a lot of still slowness in adoption in large enterprises. But it's not because people aren't conv…”
Levie: Most time in companies is spent on necessary but non-strategic work
“The vast majority of time inside of a company is on the stuff that really is not strategic. It's sort of necessary work, but it's not strategic to get done.”
Levie: In 10 years, most enterprise work will be previously uneconomic tasks
“If we go 10 years into the future, The vast majority of work that gets done in 10 years from now will be work that today is in that category. It's the work that like right now we can't even attack because we're like, I'm not going to hire somebody, pay them a …”
Levie: Box only translates ad campaigns into 3 to 5 languages due to cost
“When we, you know, launch an ad campaign internally, we translate it into like three to five languages. Our top markets, that's about all we have time for because it's just too expensive.”
Levie: Startups have few viable vectors to compete with Gusto in payroll
“You don't really want to compete with Gusto because Gusto is still a modern, really good payroll system. There's not like a lot of vectors into competing with Gusto.”
Levie: Salesforce and Marc Benioff Will Successfully Implement AI for CRM
“It won't just be like, oh, it's CRM, but with AI, because Mark Benioff in Salesforce is going to do CRM with AI. Like he's going to figure out a way to do that. He's they're very good at executing like that will happen.”
Levie: 100 Startups Created Between 2024-2028 Will Reach $5B to $20B Valuations
“There will be 100 startups that get created between last year and in three years from now that all become five, 10, twenty billion dollar companies because they're able to find the next set of nouns and verbs”
Levie: AI agents replace per-seat SaaS pricing with work volume pricing
“Agents basically completely blows that up because all of a sudden you can have AI agents that effectively contain the labor of that job function in the software itself. So you can go to a company and you can say, I know you only have three lawyers, but my agen…”
Levie: Pure consumption pricing threatens software recurring revenue predictability
“The only, you know, concern that you have to have of going of overly veering on consumption is the recurring nature of the revenue. You generally want to be in a position where you have some kind of subscription fee for your software, as opposed to only being …”
Levie: Customers will pay for workflow software rather than raw tokens
“So eventually we're going to get to a point where customers are no longer going to just be paying for the intelligence tokens. They're going to be paying for the workflow software that goes on top of the tokens themselves.”
Lieb: Google Photos operates at a 90-plus percent gross margin
“And it turns out it's like a 90 plus percent margin.”
Levie: AI tool makers will maintain flat pricing despite falling compute costs
“It's like, it's 20 bucks a month, it's 50 bucks a month, but we know that in 10 years from now, they'll probably be able to drive down their raw materials lower, but we won't be able to command lower prices for those things, because it's just like within a rea…”
Levie: B2B founders must master three essential strategy books
“Read Innovator's Dilemma.
Read Crossing the Chasm.
Read Blue Ocean Strategy.
Those are the three books you have to read.
If you do what's in those books, and you are going after the B to B market, I guarantee you will be 10 times better off than any other star…”
Levie: Do not start companies in markets untouched by AI transformation
“If your market is not truly transformed by AI, don't touch it. It's not, it's just not worth it because you're going to be fighting against a headwind that is just unnecessary to fight against. Like go after markets where AI like fundamentally changes like the…”
Levie: 20-somethings should defer existential questions and focus on commercial grind
“Like you're in your twenties. So like I was just heads down grinding, not overly worried about the meaning of life. So I'd say like, put a pin in that. Be nice to everybody. Like, help the world as much as possible. Check back in in about five years. Like, now…”
Levie: Founders should always overestimate competitors' capabilities when planning strategy
“I think it's always good to overestimate your competitor's capabilities and then figure out your strategy in a world where they have those capabilities.”
Levie: Incumbents will only sell AI agents to existing install bases
“Because I think incumbents are only going to be the agent providers for their existing install base, which means that there's going to be tons of opportunity for so many more agents that those incumbents aren't already selling into.”