Who Is Lai Jie? The Tencent Roboticist Behind Astribot

In April 2024, a video appeared of a robot doing the most ordinary things imaginable — yanking a tablecloth out from under a tray of wine glasses, pouring a glass, shaving a cucumber, flipping a sandwich in a pan, folding laundry — and the internet lost its mind. It moved with a smoothness no humanoid demo had shown before, and half the comments were some version of "there's no way this is real." The robot was the Astribot S1, and the man who built it is a quietly formidable engineer named Lai Jie (来杰), who had just walked out of one of China's most elite robotics labs to do it.

So let's get to know him — because his story is more interesting than the viral clip, and so is the honest answer to the question everyone asked about that demo. Where he came from, why he left Tencent, why his robot is strung with cables like a puppet, and whether Astribot, now a roughly $1.4-billion unicorn, has the substance to back the spectacle.

First, the short answer: who is Lai Jie?

Lai Jie is the founder and CEO of Astribot, a Shenzhen company often described as the "Chinese Figure" for trying to own both the AI brain and the body of its robot. Unlike the showmen and the outsiders elsewhere in the humanoid race, Lai is a pure roboticist — a man with roughly sixteen years in the field whose career reads like a tour of China's most serious robot programs. He's the kind of founder who didn't need to learn the technology to start a company; he was the technology.

From Baidu to Tencent's secretive robot lab

Lai (described in Chinese coverage as "post-1985," so still in his late thirties) studied at Xidian University and took a master's in intelligent systems at Wuyi University. After an early research stint, he joined Baidu in 2014 and led the team behind its Xiaodu (小度) consumer robot. Then, in 2018, he made the move that defines his résumé: he joined Tencent's Robotics X, the gaming giant's ambitious, low-profile robotics lab, as part of its founding cohort — he's frequently called the lab's "number-one employee."

At Robotics X he ran the embedded-systems group, the team responsible for the software and electronics that make a robot actually move. Under his group came Tencent's self-balancing wheeled robots and its wheel-legged machines — including, by most accounts, the acrobatic Ollie. (In the interest of honesty: independent reporting confirms he led the embedded group behind those robots; the specific "he personally led Ollie" line traces to company-seeded profiles rather than first-party Tencent material — true in spirit, worth a caveat.) Either way, this is the unglamorous heart of robotics, and Lai lived in it.

Lai Jie's whole career is a deliberate climb toward one moment: the point where AI got good enough that a robot could finally be useful in a home. He spent years building the bodies and the brains separately at Baidu and Tencent — then walked out to put them together himself.

Leaving to build a robot for the home

In late 2022, Lai left Tencent and took five or six colleagues from Robotics X with him — including co-founder Dai Yuan (戴媛), a UCLA robotics PhD — to start something of their own. On December 5, 2022, they founded Astribot in Shenzhen. The English name comes from the Latin ad astra per aspera — "through hardship to the stars" — rendered in Chinese as 星尘, "stardust." Lai's thesis was unambiguous: the fusion of AI and robotics, he says, is a "once-in-50-years opportunity," and Astribot's goal is a new generation of AI robots that can "enter every household" — a job he estimates will take five to ten years.

His core engineering conviction is a phrase worth remembering: "Design for AI." Instead of building a fixed mechanical body and bolting AI on top, he lets the demands of the AI model shape the hardware. And his founding question was almost poetic: "Why can a blind person open a door, but a robot can't?" A blind person succeeds through touch and compliant, gentle force. That insight pushed Astribot toward its most distinctive choice — rope-driven actuation.

The S1, and the demo that broke the internet

The Astribot S1 doesn't look like its rivals. It is not a walking humanoid; it's a rope/cable-driven dual-arm robot mounted on a wheeled, omnidirectional base. Those cables are the point: like tendons pulling on muscles, they let the arms move with speed and compliance that rigid, motor-in-every-joint designs struggle to match. The numbers Astribot quotes are genuinely startling — an end-effector top speed of over 10 metres per second and a 10-kilogram payload per arm, roughly a one-to-one strength-to-weight ratio.

That speed and grace is exactly what made the April 2024 demo go viral — and exactly what made people suspicious. So let me give you the honest version, because it matters more than the hype.

This is the same "demo versus reality" gap we keep flagging across the industry in how you actually test a humanoid robot. To Lai's credit, Astribot's claims were narrower and more defensible than most — real-time speed is a real, checkable thing — and the company has since shown the S1 doing autonomous work. But the lesson holds: a jaw-dropping clip is the start of the conversation, not the end of it.

The money, and the "Chinese Figure"

The capital has arrived in a hurry. Astribot raised an angel round in 2023, a Pre-A led by Matrix Partners China in 2024, and successive rounds in 2025 with Ant Group (Alibaba's affiliate) as a lead. Then, in June 2026, it announced a Series B of over 1 billion yuan that valued the company at over 10 billion yuan — roughly $1.4 billion, making it one of Shenzhen's embodied-AI unicorns.

One honest note on the cap table, since it's widely misreported: Astribot is often called "Tencent-backed," but that's really an ecosystem and talent link — the founder and core team come from Tencent Robotics X, and an early backer was founded by a Tencent co-founder. The confirmed direct strategic investor from the Big Tech world is Ant Group, not Tencent. It's a small thing, but it's the kind of detail that separates a profile from a press release. (Tencent's money, for what it's worth, went instead into AgiBot.)

What you're really buying with that valuation is the bet that rope-drive is the right road. Astribot claims to be the first to mass-produce rope-driven AI robots, and in 2026 it broadened the lineup with the T1, a cheaper wheeled humanoid starting around 89,900 yuan (~$12,500), aimed at retail, hospitality and the home. It has begun thousand-unit-scale deliveries — real traction, even if it trails the volumes of Unitree and AgiBot.

His most contrarian idea: keep the human in the loop

Here's what makes Lai genuinely interesting, and a little out of step with his peers. While everyone else promises fully autonomous robots Real Soon Now, Lai openly argues for keeping a human in the loop — at least for a while. He pitches the idea of "expert digital avatars" and makes a strikingly human case for teleoperation: an elderly person, he suggests, would often rather have "a daughter remotely operate the robot to peel an apple for them" than have a robot do it automatically. The point of the robot isn't just the chore; it's the connection.

It's a refreshing thing to hear from a robotics founder — an acknowledgment that the technology isn't ready to be left alone yet, and that the first killer app may be augmenting people rather than replacing them. He frames Astribot's market as "incremental": amplifying human capability and reaching across the distances that separate carers from the people they love, rather than simply deleting jobs.

What he gets right — and the risks

Strip away the viral clip and the case for Lai Jie — and the worry — both come into focus:

• Deep, real pedigree. Baidu and Tencent Robotics X is about as serious a robotics background as exists in China; this is an engineer's company, run by the engineer.
• A genuinely different bet. Rope-drive is a real technical fork, not a marketing gimmick — it targets the soft, compliant manipulation that rigid humanoids find hardest.
• Intellectual honesty. The human-in-the-loop, "incremental market" framing is more candid about today's limits than most of the fully-autonomous hype.
• The risks: autonomy and form factor. The questions raised by that first demo haven't fully gone away, true autonomy at home is still unproven, and a wheeled robot — while practical — may cede the long-term "humanoid" narrative to the bipedal players. At a $1.4-billion valuation built largely on promise, the shakeout we keep warning about is a real risk.

The quick facts on Lai Jie

Frequently Asked Questions