How AI Is Changing Enterprise
Aaron Levie (guest), Garry Tan (host), Jared Friedman (host), Diana Hu (host), Harj Taggar (host)
In this episode of Y Combinator, featuring Aaron Levie and Garry Tan, How AI Is Changing Enterprise explores enterprise AI Revolution: From Commodity Intelligence To Explosive Software Growth Aaron Levie of Box joins YC partners to discuss how AI is transforming enterprise software, arguing that intelligence is becoming a cheap, fungible commodity while value shifts to the workflows, integrations, and outcomes built on top of models.
Enterprise AI Revolution: From Commodity Intelligence To Explosive Software Growth
Aaron Levie of Box joins YC partners to discuss how AI is transforming enterprise software, arguing that intelligence is becoming a cheap, fungible commodity while value shifts to the workflows, integrations, and outcomes built on top of models.
He explains that enterprises do not want models per se; they want end-to-end outcomes that plug into existing systems, with model choice increasingly abstracted away and interchangeable as costs converge toward the price of raw compute.
Levie draws parallels to the cloud transition, predicting that AI will massively expand software TAM by enabling companies to automate work they never did before, rather than simply replacing existing labor budgets.
He anticipates an AI-driven abundance economy—if regulatory and societal choices cooperate—where productivity gains are reinvested into better products, more competition, and ultimately improved services and lower costs for consumers.
Key Takeaways
Don’t build ‘just a wrapper’—own a full workflow and outcome.
Levie argues that real value in B2B AI lies in the software around the model—workflows, integrations, business logic, and proprietary customer data—so startups should solve end-to-end problems rather than thinly exposing model outputs that platforms can absorb.
Get the full analysis with uListen AI
Treat models as interchangeable infrastructure; design for abstraction.
Enterprises increasingly care about outcomes, SLAs, security, and integrations more than which model is used; successful AI apps will be architected so models can be swapped as prices fall and capabilities converge, improving margins over time.
Get the full analysis with uListen AI
Pure-play model companies are a bad standalone business bet.
With OpenAI, Anthropic, hyperscalers, and open source (Meta, DeepSeek) driving prices down, Levie believes model-only businesses without a substantial software layer, distribution, or ecosystem lock-in will be squeezed by open models and cheaper incumbents.
Get the full analysis with uListen AI
Most enterprises will buy AI for ‘context’ and build AI for ‘core.’
Using Geoff Moore’s ‘core vs context’ frame, Levie says companies should buy AI-powered systems for generic functions (HR, ERP, CRM) but build proprietary AI where it directly differentiates their offering, such as trading algorithms, drug discovery, or personalization.
Get the full analysis with uListen AI
AI will expand software markets by enabling entirely new work, not just replacing labor.
Many AI use cases (e. ...
Get the full analysis with uListen AI
Usage-based and outcome-based pricing will become more common in AI SaaS.
Because AI capacity is elastic, many startups are successfully replacing BPOs or services with usage-based contracts (e. ...
Get the full analysis with uListen AI
Competitive pressure will drive reinvestment of AI efficiency gains, not permanent job elimination.
Levie contends that in competitive markets, companies will use AI-driven productivity gains to build better products and grow faster, hiring in other functions, rather than simply sitting on higher margins—pushing long-run benefits toward consumers and society.
Get the full analysis with uListen AI
Notable Quotes
“An enterprise doesn’t want a model. It wants an outcome.”
— Aaron Levie
“We used to have an API into compute and storage. Now we have an API into intelligence.”
— Aaron Levie
“The cost of intelligence is going to go to zero. It’s absolutely guaranteed.”
— Aaron Levie
“Most of the way AI will show up to a knowledge worker in 2030 will be from what we would have thought of as an ISV ten years prior.”
— Aaron Levie
“The revolution does not have to be Black Mirror. It could be something that is driven by Jevons paradox, driven by abundance for everyone.”
— Garry Tan (closing reflection on Levie’s vision)
Questions Answered in This Episode
How should a new B2B AI startup decide whether its product is dangerously close to something a model provider or ChatGPT could simply fold into their UX?
Aaron Levie of Box joins YC partners to discuss how AI is transforming enterprise software, arguing that intelligence is becoming a cheap, fungible commodity while value shifts to the workflows, integrations, and outcomes built on top of models.
Get the full analysis with uListen AI
What concrete criteria can an enterprise use to separate ‘core’ AI capabilities it must build from ‘context’ use cases it should buy off the shelf?
He explains that enterprises do not want models per se; they want end-to-end outcomes that plug into existing systems, with model choice increasingly abstracted away and interchangeable as costs converge toward the price of raw compute.
Get the full analysis with uListen AI
If model quality and pricing converge, what durable moats will remain for AI application companies beyond data and workflow integration?
Levie draws parallels to the cloud transition, predicting that AI will massively expand software TAM by enabling companies to automate work they never did before, rather than simply replacing existing labor budgets.
Get the full analysis with uListen AI
How can policymakers and city planners ensure that AI-driven productivity gains actually translate into lower real-world costs (e.g., housing, healthcare) rather than being blocked by regulation?
He anticipates an AI-driven abundance economy—if regulatory and societal choices cooperate—where productivity gains are reinvested into better products, more competition, and ultimately improved services and lower costs for consumers.
Get the full analysis with uListen AI
Given Levie’s prediction that most knowledge work software will be AI-powered ISVs, where are the biggest greenfield opportunities for startups that don’t look like traditional SaaS categories?
Get the full analysis with uListen AI
Transcript Preview
... wait a second, if we could use AI to automate more, we can build more. If we could build more, we could lower the cost of things. If we can lower the cost of things, then we can actually lift up anybody's lifestyle right now.
I think that we're in the middle of the revolution, and the revolution does not have to be, uh, Black Mirror. It could be something that is driven by Jevons paradox-
Yeah.
... driven by abundance for everyone, and that's certainly the, uh, the timeline we wanna be on, so...
That's the future I'm betting on.
Welcome back to another episode of The Light Cone. I'm Garry. This is Jared, Harj, and Diana. Uh, we're partners at YC, and collectively, we've funded companies worth hundreds of billions of dollars. And today, we have a really awesome guest, Aaron Levie-
Oh, thank you.
... of Box. (laughs)
(laughs)
He's-
Love that intro.
Yeah. (laughs)
(laughs)
Aaron, you're o- one of the best product CEOs out there.
Thank you.
Public company as well.
That's what I w- that's what I write in my Wikipedia.
Yeah, yeah. Uh, that's how I classify you. We're in the middle of the AI revolution, so, uh, how are you feeling?
Oh, pretty good.
Yeah. (laughs)
(laughs) Um, it's a, it's a good time to be in, uh, in, in software right now. Um, so yeah, feeling pretty good today.
Something we've been speaking about for a while, which I think we probably agree on is that the ChatGPT wrapper was like a bad meme, and that actually there's, like, lots of value, and always has been, in building apps on top of these foundation model companies.
In fact, the opposite might be true.
Yeah. Which is gonna be more valuable-
Yeah.
... in 10 years time, right?
Yeah, I think so. So, it's interesting. So, um, there's, there's a l- like pr- 2% truth in, in the meme and then 90% n- not truth. And so, I mean, PG, you know, with the sort of wedge theory, is like actually you do want something that is, is sort of a simple product that finds a little wedge and then you expand from there. You know, in the early days of, of cloud, if you were to be building software that let you manage documents and data, th- you would've been like, "Well, that's a wrapper on Amazon." And, and it was a total misunderstanding of the, the entire scale of software you have to build to make the storage bucket be useful in, in a particular application. So, on the, on the wrapper conversation, the exact same thing is true, which is how much software do you need around the workflow and the proprietary, you know, sort of business logic and the data that the customer brings. That's actually the value, not the, not just like what is the, the set of tokens that are coming out. Where it's a little bit true why, why startups should be, you know, at least think a couple steps ahead is you probably don't want to be something that just ChatGPT would incorporate. So it's less that the model, uh, will incorporate your value proposition, it's more if there is, you know, an a- i- if the model provider also has a consumer scale application. Like you don't want to be right in the way of something that ChatGPT will just fold in directly into its functionality. So in that case, I think you have to be, you know, sensitive to being kind of, you know, a "wrapper".
Install uListen to search the full transcript and get AI-powered insights
Get Full TranscriptGet more from every podcast
AI summaries, searchable transcripts, and fact-checking. Free forever.
Add to Chrome