Airbnb CEO Brian Chesky's latest bombshell statement: Silicon Valley's understanding of AI may be wrong.

On Airbnb's earnings day, CEO Brian Chesky sat down for a rare chat: why ChatGPT is not the way to book travel, why Silicon Valley has taken the wrong direction with AI, and why this is an era where a "one-person company" can be worth a billion dollars.

Airbnb's revenue is accelerating — it grew 10% for the full year last year, and by 17% in the first quarter alone this year. This is unusual for a platform company that has been listed for many years and has a market capitalization approaching hundreds of billions of dollars.

"For marketplace companies, accelerating growth is extremely difficult. It's like gravity: the bigger the company gets, the more the growth rate tends to drop," Chesky said. "Our US business is accelerating, and the so-called 'mature markets' like Canada, Australia, the UK and France are not actually mature at all — we still have huge room for growth."

But he believes that what will truly push the company into its "third chapter" is AI.

1. Silicon Valley Got AI's Direction Wrong

Chesky's attitude towards AI is very clear — AI is the best thing that has ever happened to Airbnb. But he raised sharp doubts about the current direction Silicon Valley is taking.

"Almost everyone in AI is building enterprise applications. Out of the last 175 companies in the latest YC batch, 159 are doing enterprise services," he said. "The enterprise market is already extremely crowded. Consumer AI is a huge untapped gap."

The data he cited is thought-provoking: although the topic of AI is extremely hot, ordinary Americans are not enthusiastic about AI. Outside Silicon Valley, the concept of AI is not well received.

"We need to build more products that ordinary people can use, so that they can say: 'I love AI because it lets me access a doctor whenever I need to,'" Chesky said. "This is not just a narrative problem — we really do need to develop more consumer-facing products."

2. ChatGPT Will Not Be The Way To Book Travel

In the travel industry, Chesky has his own judgment on the application of AI.

"ChatGPT might become a way to discover destinations, but I don't think it will be the way people book travel," he said.

Airbnb previously chose not to be a launch partner for ChatGPT's third-party plugins — Expedia and Booking did, but later ChatGPT shut all of them down anyway.

"A year from now, you will see Airbnb become very AI-native, and the search paradigm will be completely different," he said. "AI search will emerge, and we will move upstream into the destination discovery space."

He predicted that there will be three major AI transformations in the travel industry in the next 12 months: AI search, destination discovery, and customer service. But he does not believe that chatbots will become the main place for booking transactions.

3. A One-Person Company Can Be Worth A Billion Dollars

As a founder with a non-technical background, Chesky has a deep feeling about the lowered barriers to starting a business in the AI era.

"It has never been easier to start a company than now," he said. "I think a one-person company can also become a billion-dollar company, and maybe we have already seen one."

He said he never considered himself a "technical founder" before. "But today, I would be. In fact, I am coding right now. I'm using Claude Code."

His point is very clear: Coding has become a kind of language. The path from English to code is already very feasible.

"Agents can teach you to code, because you can watch them code. You can expand to look at the commands, and gradually build an intuition for 'what is possible'."

4. Starting A Company Has Never Been Cheaper, And Never Been More Expensive

Chesky pointed out a paradox in current AI entrepreneurship.

"On the one hand, it has never been cheaper and easier to start a company. On the other hand, it has never been more expensive to start a company."

Want to make a simple agent? It costs almost nothing, and anyone can do it. But if you want to do something really interesting in the AI field, you may need thousands or even tens of thousands of Blackwell GPUs — that could cost you a billion dollars, and researchers are also very expensive.

"The question is: can the YC methodology — a group of 25-year-olds with no money and no experience — compete with Google, Anthropic and OpenAI to build a trillion-dollar multi-generational company?"

"I don't know. I'm on the side of the former. I support that anyone has the ability to build a multi-generational company. But that's the crux of the debate."

5. Airbnb's "Third Chapter": From Short-Term Rentals To A Platform

Chesky divides Airbnb's history into three chapters:

Chapter 1: A crazy idea succeeded, product-market fit, ultra-high-speed growth.

Chapter 2: The pandemic — facing the abyss directly, losing 80% of the business, rebuilding the company from scratch, going public, and truly becoming a mature company.

Chapter 3: Reinventing itself — moving from being a "one-trick pony" focused on short-term rentals to a broader community, a global community built on trust that allows you to "travel and live anywhere".

"That means homes, hotels, services, experiences, and many more business line expansions," he said. "It's a bit like Amazon moving from books to everything, or Apple moving from Mac to iPod — that's the next wave of growth."

6. The Road To Super App: One Year To One And A Half Years

Chesky has been talking about the concept of "super app" since the early days. His judgment is: Airbnb is still one to one and a half years away from that goal.

"First of all, it's not certain whether a super app can emerge in the United States. They exist in China, the most famous one is WeChat. Maybe Apple's App Store is a super app. But can there be a second one? I don't know."

But he is very sure that in the next 12 to 18 months, you will see "some very big" changes in Airbnb.

For every business you have already seen, there are two or three, four or five more businesses we haven't talked about that are being incubated.

When asked what can bring Airbnb closer to the super app vision, he mentioned several key words: identity verification, preference storage, payment platform.

"We built our payment platform before Stripe existed. We have more than 200 million verified internet identities. In the AI era, you want to know if the people on the internet are real — this is a problem that can be solved, and it is a useful solution."

7. AI Agents: We Are Just Getting Started

Chesky confessed that his position on AI agents is: "I believe in agents, but I have reservations."

"I don't think the way chatbots are currently built, or even agentic programming agents, is the solution to all problems."

He made an analogy: "If you double something 10 times in 10 years, that's 1000 times. I think we are at the starting point of a thousand-year journey for AI."

In his view, Claude Code is great, ChatGPT is useful, but most people are still using it as an "interactive search engine". "We have only scratched the surface of exploring the possibilities of AI."

He was surprised by two things: first, AI is far more powerful than he imagined, saving him a lot of time and making him more efficient; second, getting large companies to change their workflows is far more difficult than expected.

"Large companies have inertia, a resistance to change. But if you can overcome that hurdle, it's like swimming in warm water afterwards."

8. The Revival Of Consumer AI Is Coming

Chesky predicted: In the next two to three years, consumer AI will usher in a revival and begin to truly change people's daily lives.

"The internet changed our daily lives, the iPhone changed our daily lives. When will AI change our daily lives? I'm not sure yet — it doesn't necessarily have to be us driving this, we will be part of it — but my prediction is that you will see a revival in the consumer AI space in the next two or three years."

He believes that the current penetration rate of AI on the consumer side is very low: little progress in video, little progress in photos, almost no progress in new native AI interfaces, and voice commands are far from being popularized — "Most people are talking, but they are still typing on their computers."

"So I still think it's extremely early right now."

Conclusion

When asked what he would do if he started a company from scratch today, Chesky said: "I would code. In fact, I am coding right now."

His core message is clear: AI is not a threat, but an opportunity. But the opportunity does not belong to those who only build enterprise applications, those who only treat AI as a search tool, nor those large companies that are afraid of change.

"This is one of the four most special periods in the history of computing, and it may be bigger than the other three combined."

And for founders, the best era has arrived — "a one-person company can also be worth a billion dollars" — but the premise is that you dare to find a breakthrough in the consumer space, instead of crowding into the enterprise market to compete.

This article is from the WeChat official account "AIGC Index", author: Mark, published with authorization from 36Kr.