Vinod Khosla is more popular now than ever. The co-founder of Sun Microsystems has turned into a prominent investor, first at Kleiner Perkins and for the past two decades at his venture firm Khosla Ventures. Thanks to his sincere advice and the company's track record of following, it has always attracted the attention of founders. Bet on Stripe, Square, Affirm, and DoorDash. But a $50 million gamble on OpenAI in 2019 – when it was unclear whether the company would succeed at its current scale – put Khosla Ventures and Khosla himself squarely in the spotlight.
He's having a really good time. I sat down with Khosla last week at the Collision conference in Toronto. Before going on stage, he said he's been making public appearances several times a week lately, whether on stage, on podcasts or in TV interviews. For example, when asked if he was exhausted from his schedule because he had flown to Toronto just hours before we sat down, he shrugged off the suggestion.
There are certainly things he likes to talk about, and the art of closing a deal is not one of them. “Honestly, the investor side is much less interesting to me.” When I asked him about what I had recently heard, he said: The thing is that since he started Khosla Ventures he hasn't taken a single dollar in management fees, even though he currently earns a management fee of $18. We manage $1 billion in assets. (I confirmed this, but stated that this only applies to me and is not a company-wide policy.)
We spoke about this gap because he is even more passionate about the startup opportunities he is spying on in an environment that is changing every day with advances in AI. We also talked about what worries him most about the ramifications of AI. FTC Chairman Lina Khan; And in his view, “Europeans have regulated themselves from being leaders in any field of technology.”
We first talked about Apple's flashy new deal with OpenAI that will allow Apple to integrate ChatGPT into Siri and its generative AI tools. Apple could strike similar deals with other AI models, including Meta, but naturally, as an OpenAI investor, Khosla is optimistic about the only partnership Apple has publicly announced so far.
Khosla called this a “validation” for OpenAI. By announcing the deal with OpenAI at a high-profile developer conference, Apple said it was “expressing confidence that (OpenAI CEO) Sam (Altman) will lead (AI development) over the next five to 10 years.” Khosla. “When a company like Apple invests in a technology, they typically don’t change it the following year.”
But is that both good and bad news for Khosla? As TechCrunch notes, many startups are likely to collapse due to some of Apple's latest features, and Khosla's portfolio companies don't seem to be entirely immune. I was particularly curious about Rabbit, an AI-based hardware device that promises to act as a sort of AI assistant for users and is backed by Khosla Ventures.
When asked if Apple could retire the device, Khosla suggested the device is more flexible than people imagine and could eventually be used in businesses like hospitals, including emergency room settings. He put it into more and more things that “watch what you do, see what you do, and automatically react.”
In fact, Khosla suggested that as large-scale language models like OpenAI become more advanced, his team actively avoids anything that could become “roadkill.” And he highlighted at least one company that is not in his portfolio. Grammarly is a writing assistant startup that was recently valued by its backers at $13 billion.
“For example, if you're doing Grammarly, it's a simple wrapper around today's model and Grammarly won't be able to keep up. It wasn't supposed to be an app. It shows the need for that feature, but it will be part of Word or Google Docs. It's pretty clear. When we talk to YC companies or other companies, I can usually say, ‘Half of these companies are going to be obsolete before the end of their YC deployment,’” Khosla continued.
Khosla believes there are many opportunities in an industry where expertise is available almost for free. But it's not clear (after asking him) how these companies continue to make money. Think tutoring or oncology.
Khosla said: “Open AI or Google will not create chip designers (for smartphones). OpenAI and Google will not train structural engineers. “They’re not going to train primary care physicians or mental health therapists,” he said. “So there are so many areas that (starters) can do. But they need to look at where the model is going to be next year and five years from now and say, 'We want to leverage that capacity.'
We also talked about regulation. I've seen Khosla say before that we need to protect large, closed language models like OpenAI. Although there has to be a regulatory framework around it. I wondered if this meant that Khosla would forever reject other “open source” AI.
He said no, noting that he was a “big fan” of open source. Sun was one of the first companies to “jump into open source,” he said, releasing the source code for its file system. He also noted that Khosla Ventures was the original investor in GitLab, software that invites people to work on code collaboratively.
But he suggested that in the context of large-scale language models, open source is a completely different animal. “The biggest risk we face when it comes to AI is China” and “strong Chinese AI” competing with America’s “liberal values,” he said, adding, “We need to make sure China stays behind us.” Otherwise, he warned, China will provide 'free doctors and free oncologists' to the world, and while they do so, they will 'export both the economic power and political philosophy that comes with AI.' ”
On stage, I mentioned to Khosla my recent sit-down with FTC Chair Lina Khan, who does not believe in the national champion model as a reason why companies like Google and OpenAI are pushing AI development.
Khan always hears from executives and investors who say government intervention will lead America down a dangerous path. But during our sit-down with her, she argued that the United States had chosen “the path of competition,” which “ultimately fueled and fostered many of these groundbreaking innovations and many of our incredible growths.” The country had fun and it helped us get ahead globally.”
If you look at other countries that have instead chosen the national champion model, they are the ones left behind,” Khan added at the time.
But I barely mentioned Khan when Khosla dismissed her, calling her “not a reasonable human being” and accusing her of not understanding business.
“She shouldn’t be playing that role,” Khosla said. “Antitrust laws are a good thing in any country and any economic system. But over-enforcing or over-enforcing antitrust laws is bad economic policy. One of the advantages the United States has over its European competitors is its much more reasonable business environment. This is why Europeans have been restricted from taking a leading role in any field of technology. They basically regulated themselves in AI, all social media, all internet startups.”
Of course, if some antitrust enforcement is good but too much enforcement is not, the question is where to draw the line. At this point, before we part ways, I mentioned the ‘richness’ that Altman predicts will be created by AI. During TechCrunch's StrictlyVC event last year, Altman said the “good practices” for AI were “so unbelievably good that when you start talking about it, you sound like a crazy person.”
Khosla says he feels the same way, but he's long wondered how exactly society will reap all these positive benefits if regulators don't become more involved in the trajectories of these companies. In the end, I spoke to Khosla on stage. We have already seen vast amounts of wealth and power tied up in increasingly smaller groups of companies and individuals. When will enough be enough?
Here Khosla said this issue troubles him greatly. “I think I’ll still be working 25 years from now. . . The need to work will largely disappear.” Still, he said AI should create “great abundance, big GDP growth, big productivity,” “all the things economists measure,” but he worries “more than anything about the growing income gap.” “How can we distribute the benefits of AI fairly?”
He has an idea where the turning point will be. “If (US) GDP growth falls from the current 2% (currently less than 1% in Europe) to 4%, 5%, 6%, we will have enough abundance to share the wealth and share the benefits.”
Of course, whether and how that will happen is a bigger question, and one that Khosla, a self-proclaimed techno-optimist, has no answer to despite his extraordinary talent.
Instead, after giving his final closing remarks, he thanked the crowd for their time and stood up to leave the stage. There, a dozen young founders gathered side by side, hoping to keep his ear as long as possible.
