As a partner in Khosla ventures, Ethan Choi he is not ashamed to say what he thinks.
The investor has been vocal about its belief that AI poses a huge threat to entry-level jobs, or what it sees as a changing social contract in today’s workforce.
He would know about the impact of artificial intelligence. Choi led investments in several corporations focusing on artificial intelligence.
Known for his founding credentials, he also backs enterprise software and fintech infrastructure corporations.
Before joining Khosla in 2024, Choi was a partner at the firm Speed upwhere he led and managed significant development investments in corporations corresponding to 1Password, Klaviyo, Letter (which was taken over by Visa), Nuvem Store AND tools of the trade. Before Accel, he worked at Ghost capitalsupporting corporations corresponding to Lynda.com (acquired by LinkedIn), Headroom, Bright software AND PicMonkey (acquired by Shutterstock).
I recently spoke with Choi to learn more about why he thinks entry-level jobs could also be disappearing, why he modified his investing philosophy, and how he went from growth investing to being stage agnostic.
This interview has been edited for brevity and clarity.
Crunchbase News: You’ve had great results trading over the last few years Ramp, ClickHome, Vercel, Collect, Bridge and others. How do you manage this volume?
Choi: It was an intense episode. There were about seven deals last 12 months alone, which was a crazy 12 months by any measure. This 12 months has been a bit quieter as I focus on settling into the corporations I have invested in.
You are currently investigating the disappearance of entry-level jobs. As a parent, this sounds a little scary to me. What do you see?
Artificial intelligence is a huge puzzle. I see this in my very own work process. I exploit models to know technical capabilities – asking about Clickhouse’s indexing methodology in comparison with it Snowflakeit’s in voice mode when I’m driving. He draws from scientific articles and documentation faster than any human. I describe it as feeling like I’m wearing an “Ironman suit.”
The problem is that if I can do the work of a junior associate on my very own, almost immediately, those roles disappear. We face a world where the essential work we used to rely on young people to do is now at stake.
If the era of on-the-job training involves an end, where will students and universities be left?
The burden of the first three years of “learning to be a professional” should be transferred to universities. I look at the traditional American model of general education requirements and think, “Why do we do this?” We did it in highschool. Universities ought to be places where AI is used to really create things and apply knowledge in the real world.
If you are a computer science major today, you must graduate looking and performing like a third- or fourth-year engineer. The bar has been raised for everyone. Although schools prefer it Vanderbilt AND Arizona State University rely on an “AI-first” curriculum, many elite institutions remain silent as they fight to work out methods to adapt.
Khosla is known for her contradiction. How does this translate into the growth stage in such a competitive market?
I’ve actually evolved and I’m a stage agnostic. While people keep putting me in the “growth” bucket, I’m doing a lot more seed and Series A. In this era, a company’s metrics today are no guarantee of where it can be in two years because the rate of change is so high.
I modified my philosophy: it was 80% metrics and 20% founders. Now 90% of them are founders. The only constant is how unique the founding team is and how quickly they will adapt. If code is written 10 times faster, a company could face 50 years of change in one decade. We have to support individuals who can cope with this stress.
You predicted “mass slaughter” for some software corporations. Who will survive the transition to a world based on artificial intelligence?
We are moving from trading on revenue multiples to trading on free money flow and PE multiples. It’s a painful change. The market now must imagine that the company is born on artificial intelligence – that its revenues are moving towards models based on inference and usage, not only old seat licenses.
I expect a carnage for lightweight, horizontal applications and mid-market corporations unable to draw AI talent. I have a lot of respect for founders like those from Intercom Or Air table who “burn the boats” to reinvent their entire business. It’s extremely difficult, but in this market you either reinvent something or get replaced by someone who builds natively from day one.
Given your experience in fintech infrastructure, where do you see the next “unconventional” opportunity in financial services that almost all growth investors are currently overlooking?
There is now some degree of consensus, but I still think it’s quite unconventional that registration systems might be torn up, whether in financial services or other categories.
For example, we recently invested in Double entrywho seeks to interchange NetSuite and the core accounting system, which is the last record system I believed could be at risk. We see that with AI, startups can build migration paths that did not exist before, in addition to the depth and breadth of existing platforms, in a fraction of the time.
Vinod Khosla he often talks about “challenging the conventional wisdom” of the founders. Can you share an example of a time when you needed to steer a growth-stage founder away from a “safe” path toward a much broader, albeit riskier, vision?
Generally speaking, a founder will often consider a very dangerous but potentially game-changing product addition or acquisition. While I see part of our job as investors and board members as helping them discover and manage potential risks, the most significant thing we are able to do is give them the courage to take risks that may impact the company and the category they are in.
You have noticed that talent density is the most significant variable for success. In a market where artificial intelligence is automating routine work, how have the criteria for what defines an “elite” executive have modified?
Perhaps paradoxically, one of the principal differences in the criteria is whether that executive himself has an “IC” (individual contributor) and can do most of the work required out of the gate with his own two hands and AI.