The limelight has passed, the upsurge is receding, the law of value regulates the market, and the effects of life, illness, and death begin to appear. This is the state of development of the AI industry in 2019. And the reality is cruel, and commercial, landing, and large-scale revenue are becoming the core standards for testing AI companies.
So compared to the past, in this year, the early hidden dangers will become the fuse, and small problems in business may be fatal. At least the following 10 stars AI startups, at first, no one expected that death would come so quietly and quickly.
Roadstar.ai: Financing US $ 128 million, died in Series A, internal conflict
Roadstar.ai (Shenzhen Xingxing Technology) was founded in May 2017.
Once a shining star driverless company, it has raised the US $ 128 million in funding.
The fate of failing to fail in 2019 became the first unmanned car company to fall.
There are three founders of Roadstar: Tong Xianqiao (CEO), Measurement (CTO), and Zhou Guang (Chief Robot Expert). The three belonged to the Baidu Silicon Valley driverless car team in 2016.
But it is said that their shares are too average, and no one has full control.
And Neiyi Gong began on January 21, 2019.
At that time, its two founders, Xian Xianqiao and Gao, promoted the company’s announcement and fired another co-founder, Zhou Guang, who was also representing RoadStar in Tokyo.
The announcement listed Zhou Guang’s three major counts, including long-term non-compliance with the company’s internal code management rules and regulations, deliberately falsified data in government regulatory reports, and taking the lead in disrupting the company’s internal financial rules and regulations.
After the internal dispute was disclosed, the investors involved in the coordination, and eventually the shareholders forced the liquidation to exit.
The once-starring self-driving car company became a meteor.
Jibo: raised $ 72.7 million, died in Series B, competition
Founded in 2012, Jibo is a team led by MIT Dr. Cynthia Breazeal and is the originator of home social robots.
In 2014, Jibo also raised $ 3.6 million in crowdfunding. Six successive financings have raised 72.7 million US dollars.
Shipped to the public in 2017 for $ 899.
As early as the end of last year, Jibo sent bad news: it sold its intellectual property assets and IP to New York investment company SQN Venture Partners.
In early March of this year, Jibo announced that it would soon shut down its servers:
Existing robots will continue to respond to “Hey, Jibo” voice commands, but will not be able to understand other voice commands.
Jibo’s failure stems from its many restrictions and competition from products like Alexa.
Aria Insights: $ 39 million in funding, died in Series B, and funds flow of the chain
The drone company was originally called CyPhy and was founded in 2008 by Helen Greiner.
Seven successive rounds of financing have raised a total of 39 million US dollars.
The goal is to develop a drone that is connected to a grounded power source and supports the drone to complete several days of missions for observation and communication.
But in an interview, she said that investors put pressure on her to start commercial sales before preparations were completed.
So she resigned as CEO and left the company at the end of 2018 to work as a consultant for drones in the US military.
Earlier this year, CyPhy announced that it would be renamed Aria Insights. And from the development of hardware to the development of UAV data analysis software.
In March of this year, the company closed its doors after exhausting 39 million US dollars of venture capital.
Former CEO Helen Greiner stated:
Whether it is a supply chain issue or a regulatory issue, it will hinder the rapid development of startups.
Seven Dreamers: raised 95 million US dollars, died in round B, capital flow cut off
Seven Dreamers, founded in 2014, is a Japanese robotics startup that has developed the world’s first stacking robot, Laundroid.
3 successive financings, raising a total of USD 95 million.
The goal of this robot is to be the ultimate wardrobe manager for the entire family.
It has multiple cameras to scan and uses a robotic arm to fold clothes.
It also uses AI technology to analyze and find the best way to fold each garment.
But the efficiency is relatively low, it takes 5-10 minutes to fold a piece of clothing.
In addition to folding clothes, this robot can also do laundry.
However, on April 23 this year, Seven Dreamers filed for bankruptcy with the Tokyo District Court and decided to begin relevant procedures.
Seven Dreamers can be described as difficult on the development road, but in the end, due to the problem of capital operation, it failed to persist to the end.
According to data released by the Japanese Empire DATABANK research company, Seven Dreamers has a total debt of 2.25 billion yen (about 130 million yuan).
Anki: Raised 182 million US dollars, died in Series D and broke the capital flow
At the beginning of this year, Anki, named by the Fast Company as the “2019 Most Innovative Enterprise in the World” in the field of robotics, suddenly died at the end of April.
Founded in 2010, Anki is a star startup that was highly sought after in Silicon Valley.
Financings obtained at that time amounted to USD 182 million, with a total of 5 financings.
The first product they withdrew from in 2013 was listed on Apple WWDC and was highly praised by Cook. After that, it was even rated by the media as a company that can compete with “Google”; Microsoft and Amazon both wanted to earn revenue.
Its small robot Cozmo, launched in 2016, is leading a wave of enthusiasm, becoming the officially certified teaching robot of Carnegie Mellon University.
On the morning of April 29th, local time, Anki held a staff meeting. It’s Boris Sofman told more than 200 employees that they will be fired on Wednesday and each will receive a weekly severance payment.
The employees were also told that Anki intends to look for additional funding after a new funding round fails.
In the end, financing was unsuccessful, and acquisitions by giants such as Microsoft, Amazon, and Comcast were not finalized.
So, after Anki sold more than 6.5 million products and had an annual revenue of nearly $ 100 million that’s it.
Drive.ai: $ 77 million in financing, died in Series B, management
This is a self-driving car startup that was valued at the US $ 200 million, and the Wu and Das stood on the platform.
Drive.ai was founded in 2015. The founding team is Stanford University’s machine learning researchers, including CEO Sameep Tandon and co-founder Wang Tao are doctoral students of Wu Enda, and founding president Carol Reiley is the wife of Wu Enda.
Drive.ai filed a notice to the state government in June this year stating that it was about to shut down and permanently cease operations, firing 90 employees, including the CEO.
In more than two years, it has raised 5 times in total and raised more than 77 million US dollars. The investors are Northern Light Venture Capital, Ena Investment (NEA), Jiyuan Capital, Grab, etc., and the valuation has reached 200 million US dollars.
In February 2017, Drive.ai also demonstrated the technical strength for the first time to the outside world: it released a video of a self-driving car walking through Mountain View in the rain, making a sensation on both sides of the Pacific Ocean, and also the world’s first unmanned startup company to display Yeyu driving For a moment, there was nothing in the limelight.
In September 2017, Drive.ai announced that it would provide unmanned taxis for Lyft, but there was not much follow-up news.
At the same time, Drive.ai also stated that it plans to open an office in Singapore and intends to provide services, but has not since said.
Finally, the Drive.ai driverless vehicle landed, appearing in Texas in May 2018.
However, it never occurred to us that the situation has changed drastically.
In March of this year, the foreign media The Information revealed that Drive.ai is seeking to sell itself for $ 200 million.
By June of this year, the submissions indicated shutdown.
According to the resigning employee, the finger is directed at the management, “too greedy”, and the speaker believes that the valuation is not high enough and he is unwilling to answer, which has also become the core reason for the founding CEO to be replaced.
At that time, the team began to disperse, and the army’s heart was lost, and there was no possibility of a headwind overturn.
Apple has admitted to the media that it has acquired Drive.ai, but only some engineers can wear Apple badges.
Reach Robotics: Raised $ 7.8 million, died in Series A, competition
Reach Robotics is a consumer robotics startup founded in 2013. It is the world’s first AR game robot developer.
A total of 8 financings have been raised, raising US $ 7.8 million.
The company has launched the world’s first game robot that seamlessly integrates AR technology, launched in multiple markets, and developed supporting educational products for it.
In September of this year, its co-founder Silas Adekunle posted on LinkedIn that:
Reach Robotics, a consumer robotics company based in Bristol, UK, has closed its consumer robotics business.
Subsequently, the official also responded to this:
It has been decided to close the business with effect from September 2, 2009.
The reason for the failure is simple. The consumer robotics market has been fiercely competitive and challenging. This is particularly evident in startups.
Daqri: raised $ 275 million, died in Series B, competition
Founded in 2010, Daqri is a top player in AR headsets and a former rival of Magic Leap.
Two successive private equity rounds, raising $ 275 million.
In 2014, a helmet product for the AR display was launched. At the same time, some AR-related technologies have also been transferred to the fields of automotive HUD, 3D printing and medical imaging.
It can be regarded as quite good results.
As early as last year, news of “repeated layoffs” came out.
A former employee who has witnessed Daqri’s crazy expansion from 12 to 300 people wrote on Hacker News that Daqri recruited a lot of outstanding talents, but there were huge problems with internal communication.
“When the Los Angeles team built an AR headset based on Android, all visual scientists in San Francisco were working on iOS, and all those great technologies were useless.”
Another user who has worked for Magic Leap said that in his opinion AR glasses are far from mature.
AR needs to interact with the real world. Object detection can be done through machine learning models, but this will cause delays. And this delay can make the brain feel nauseous and vomiting.
Furthermore, both AR and VR have encountered different degrees of coldness in the past two years. They also have to fight against rivals such as Microsoft HoloLens and Google Glass. It is also estimated that funding is also an important issue.
In September, several former employees and sources said that Daqri had closed its headquarters, laid off a large number of employees, and was selling related assets.
Carbon Robotics: Raises $ 5 million, dies in seed round, funds flow off-chain
Carbon Robotics was founded in 2014 and produces the world’s cheapest industrial robotic arm: another $ 20,000, and his KATIA $ 2,000-only one-tenth, the ultimate cost-effective.
In just 2 years, he was selected into the annual authoritative list of RBR50 2016 by the Robot Business Review, becoming the top 50 global robotics technology. At the time, most of the names were giants, including ABB, Fanuc, DJI, and Google, Amazon …
△ Rosanna Myers, founder of Carbon Robotics
And it is very rich in functions, including 3D printing, laser cutting, decorating cakes and so on.
You can train a skill in 30 minutes.
On October 2 this year, the board of directors chose to bring the company into bankruptcy proceedings. They chose a form of bankruptcy allowed by California law: a third party oversees the sale of the company’s remaining assets.
A draft document shows that five KATIA robotic arms, software source code, testing equipment, and a series of patents that are completed and running are all sold.
As for why this is the case, the document also mentions the immediate reason: no money.
For most of 2019, Carbon Robotics is getting tighter on hand.
If you choose to go bankrupt, you probably don’t have enough funds to support it.
Although I don’t know the specific sales and profit situation, from the perspective of financing alone, Carbon Robotics has not made enough financing reserves since its establishment.
Public information shows that since its establishment in 2014, the robot arm company has raised a total of 5.36 million US dollars, or about 37 million yuan.
In addition to the seed round of more than US $ 3 million in 2016, an additional US $ 2 million was added in 2018, but there has been no investment since.
I .am +: Financing $ 123 million, died of capital flow off-chain
I .am + Technology was established in 2012.
Maybe you don’t know much about it, but most people know its boss-Will.i.am.
The smart home platform Wink was acquired by I .am + in 2017.
Earlier, two current employees of Wink said that workers had not been paid for 7 weeks, and their studio in Schenectady, New York had been temporarily closed.
Its users have also posted statements on platforms such as Twitter, Reddit and Facebook, saying that various third-party devices have stopped working with the platform and that the company’s customer support lines have been paralyzed.
I .am + was initially positioned as a manufacturer of luxury iPhone cases, and then expanded to other hardware areas.
The company eventually raised $ 123 million in venture capital, and its business once again turned to make conversational artificial intelligence for businesses.
But several emails from all employees showed that the company’s stalled deal with Dubai-based retail giant Majid Al Futtaim put the company in trouble.
At the same time, according to documents submitted to the Secretary of State of California, I .am + also has a late payment fee for state and federal government tax payments.
The federal government applied for a lien on I .am + in August, and documents show that I .am + has arrears of taxes and interest of $ 1,787,726.11.
It can be described as the collapse of the subsidiary, the parent company is facing bankruptcy.
Why did it fall this year?
Judging from the failure of artificial intelligence startups this year, it is mainly due to insufficient funds, market competition and internal problems.
This also clearly reflects the main reason for the failure of artificial intelligence startups.
Grouping of startup projects
Jibo, the originator of family social robots, is a good example. In the market competition, it can not match similar products such as Alexa but has the fate of being eliminated.
Not only that, but more mature technologies such as face recognition and drones have made entrepreneurs swarm up and failed to find the value of their subdivisions, which makes product differentiation not obvious.
Not paying enough attention to commercialization and product landing
Many AI startups are too focused on technological breakthroughs and continue to burn money in scientific research, but ignore product landing and sales.
If the technology and the market are severely disconnected, the problem of a broken capital chain is prone to occur.
Unreasonable start-up team structure
For artificial intelligence startups, technology is undoubtedly very important.
But if there is no reasonable mechanism for team management or a direct break between leaders, a blow to a startup may be fatal.
But in a nutshell, here and now, the most critical internal cause is AI entrepreneurship and development, and it is time for delivery.
No matter how beautiful the idea is, it must withstand the test of reality, noisy investment and financing also need a day to realize revenue, and noble technology also needs to be commercialized …
Once you can’t afford the inspection at the time of delivery, the previously undisclosed problems may be exposed, intensified, and eventually fatal.
What’s sadder is that this is still a company that has been noticeably shining and attracted much attention. They have more financing, visibility, and appeal, but they still cannot escape …
And more startups outside the halo, from scratch, from 0 to 1, from life to death, maybe more difficult and unknown.
So for whom does the death knell ring?
More than these 10-star companies.
The rational return has begun, value revaluation has begun, and the upsurge has begun …
The AI startup death list may have just begun.