qixing: Atlas Demonstrates How Deep Technology Survives

In this year's global AI robot boom, the humanoid robot Atlas is undoubtedly the brightest star. However, the American robotics company behind it, Boston Dynamics, has never achieved sustained profitability since its founding in 1992, and has repeatedly faced capital impairment issues. Even after Hyundai Motor Group acquired it for approximately $1.1 billion in 2021, the company's accumulated losses over the past four years have still exceeded 1.2 trillion won.
If this company had been based in South Korea, it would likely have been labeled a "zombie company" and eliminated from the market within five years. It's hard to imagine that the capital market would be willing to wait 34 years for a technology to fully mature. Atlas's fluid joint movements on the stage at the Consumer Electronics Show 2026 last month were the result of decades of continuous investment and patient waiting. In the United States, how exactly do these seemingly "irrational" hardcore technology companies survive?
Boston Dynamics originated as a spin-off from a laboratory at MIT. In 1992, Mark Raibert, then a professor at MIT, spun off his "Leg Labs" into a company. In an era when robots were largely science fiction, the company's development was entirely reliant on government funding.
This support went far beyond a one-time startup grant. The Defense Advanced Research Projects Agency (DARPA) became the company's first client, assigning it a series of demanding research tasks. Behind these orders were highly forward-thinking ideas—such as developing transport robots to replace soldiers in dangerous terrain.
DARPA is widely recognized as the cradle of core American technology, investing huge sums of money in projects with a high probability of failure. The origins of today's mainstream technologies, such as the internet and voice assistants, can be traced back to this agency. In a 2016 white paper, DARPA officials wrote: "If none of our projects fail, it means we haven't taken enough risks." For a long time, the agency has maintained a tolerant attitude towards project failure rates as high as 90%.
It was with DARPA's support that Boston Dynamics was able to set aside the pressure of short-term profits and focus on 20 years of long-term technological accumulation. The quadruped robot "BigDog," launched in 2005, and Atlas, the first bipedal humanoid robot, launched in 2013, both originated from these long-term research projects. The US Congress has never criticized DARPA for the lack of immediate commercial results from these projects.
Entering the 2010s, robotics technology began to attract investment from major technology companies. Although the robotics market was still in its infancy, large corporations increasingly viewed it as a future growth engine. Google's $500 million acquisition of Boston Dynamics in 2013 vividly illustrates this trend and created a "capital greenhouse" for the company.
In 2017, Boston Dynamics was sold to SoftBank Vision Fund, and finally, in 2021, it came under the umbrella of Hyundai Motor Group, thus realizing the possibility of synergistic development with the manufacturing industry. This 34-year development process has been a convergence of multiple forces: universities, represented by MIT, sowed the seeds of technology; DARPA provided the nourishment for its growth; and global capital acted as a greenhouse, turning imagined technologies into reality. Now, Boston Dynamics' technology is finally seeing the dawn of mass production.
The situation in South Korea presents a stark contrast. In South Korea, it is extremely difficult for university laboratories to spawn businesses, and the government struggles to sustain funding for long-term research projects that have not yet yielded significant results. Acquisitions of startups by large corporations are extremely rare: only about 3% to 4% of startups in South Korea exit through mergers and acquisitions, with initial public offerings (IPOs) being almost the only viable way for companies to exit capital; in the United States, nearly 90% of startups are eventually acquired.
Regulatory barriers are a key reason why large-scale corporate mergers and acquisitions are difficult to pursue in South Korea. When large South Korean conglomerates acquire local startups, they face a series of stringent regulations related to the designation of subsidiaries; if post-acquisition performance fails to meet expectations, corporate executives may be charged with breach of fiduciary duty, and the resale market for these acquired companies is also very limited. Against this backdrop, the domestic startup ecosystem is already lagging behind, making the creation of a hardcore technology startup almost a pipe dream.
The Lee Jae-myung administration in South Korea recently launched a startup selection program and declared its intention to build South Korea into an entrepreneurial society. This move is undoubtedly a positive step towards fostering an entrepreneurial boom, but we should not overlook the fact that the growth of world-class hardcore technology has never relied on glamorous events, but rather on patient capital willing to tolerate failure, and the institutional support system that makes this patience possible.

looking for M22473-GRT-FRR and XM1202, Marscomponents.com independent distributor have more than 650,000 parts you can choose.