Rabbit Co-Founder and CEO Jesse Lyu Embraces Startup Survival Amid Competition
Jesse Lyu, co-founder and CEO of Rabbit, is not intimidated by the idea of failure—specifically, the potential demise of his company. He expressed how the unpredictable nature of startups can be influenced by the actions of giant competitors like Google, Microsoft, or Apple, but insists this reality doesn't warrant throwing in the towel.
Speaking at StrictlyVC LA, Lyu shared his philosophical outlook on the looming threat posed by tech titans. (Quotes have been refined for clarity.)
Rabbit’s r1, a groundbreaking pocket AI assistant, gained significant attention during its launch at CES. This compact device, half the size of a traditional smartphone, functions exclusively as a voice-powered assistant. It can not only answer questions but also perform complex tasks and manage your apps similarly to ChatGPT. Lyu emphasizes that this innovation comprises two essential components: “intent” and “action.”
“I envisioned this concept a decade ago, but the technology wasn’t ready back then. Today marks a pivotal moment in history where a device like this is truly feasible,” Lyu explained.
Lyu's curiosity about the capabilities of large language models (LLMs) to comprehend language and intent led him to explore how these systems could execute actions.
Will Rabbit's r1, with its eye-catching design, distinguish itself amidst a crowded field of virtual assistants?
“We experimented with advanced prompts to push the language model to act, but the outcomes were disappointing," he recalled. “For instance, a demo from another company attempted to use an LLM to post a comment on MrBeast’s latest YouTube video. Theoretically possible, but it required constant supervision, taking two to three minutes for a single task. Such a process doesn't yield a satisfactory user experience.”
The solution lies in their “large action model,” a sophisticated system trained on extensive user interactions with popular applications. “We analyzed hours of real-world engagements with apps like Spotify, Uber, Expedia, and DoorDash—essentially the top 800 apps. By employing a neural symbolic network, we ask the AI now dubbed the large action model to review these interactions frame by frame, enabling it to identify buttons and elements. This allows us to develop logical automation,” Lyu detailed.
While the language processing remains reliant on third-party LLMs like Perplexity—who is leveraging Rabbit’s success by providing a year of complimentary service alongside the r1—I raised concerns about potential API costs threatening the company's financial stability.
“Firstly, we’re selling the r1 without incurring losses, which is an impressive feat for a first-generation startup,” Lyu stated, crediting his talented hardware team for negotiating component costs effectively. “We’re nearing 100,000 orders. Just two days before our keynote, I hoped we would sell 500 units on launch day, but we ended up selling 18,000.”
Lyu does not foresee a subscription model being viable, particularly since the core philosophy of the r1 revolves around affordability and simplicity. However, users will have the opportunity to create and sell customized app-specific models in the future, with Rabbit taking a share—though he stressed this is a long-term goal without concrete details yet available.
Finally, when discussing the immense financial power of major corporations investing billions in AI, Lyu offered a surprisingly tranquil perspective on potentially being overshadowed by tech giants like Google, Microsoft, and Apple, especially in light of recent statements from Apple CEO Tim Cook indicating significant AI advancements this year.
“I’m realistic about our status as a startup,” he acknowledged. “The first lesson I learned from Y Combinator is that 99% of startups fail. If you think your genius idea guarantees success no matter what these tech giants do, you’re mistaken. Startups are fundamentally a survival game, and it's crucial to focus on your own journey.”
“They will pursue their path, while I pursue mine. Some founders might have quit after hearing about Apple entering the car market—now what? This level of competition can drive us to innovate faster or fail sooner; that's just how startups operate. I'm giving it my all—because ultimately, it’s a survival game.”
You can watch the full panel discussion below.
Ranking
