Rubin, the father of Android, is on a new journey: artificial intelligence

"Father of Android" Rubin Tencent Technology News February 11th, "Wired" magazine published an article recently said that after the opening of Android, the smartphone business has lost to Andy Rubin (Andy Rubin) Attraction, this is why he left from Google (microblogging). Today, Rubin, who is dissatisfied with the status quo of the world, is on a new journey: artificial intelligence.
A few years ago, Rubin, the father of Android, helped his wife Rie to open a bakery in Los Altos, California, named "Voyageur du Temps", translated as "Time Traveler." As the name suggests, the bakery allows customers to return to the past by carefully recreating the taste and taste of European classic pastries. To achieve this goal, Rubin hired a chef from Japan (because Japan has inherited traditional baking techniques well) and bought the already rare Bongard Cervap oven.
The bakery is a typical "Rubin style", with a certain amount of money, energy and engineering talent for hobbies, just for fun. But this investment in the bakery also has an "atypical" side, because Rubin usually invests in some future technologies, such as robotic arms and retina scanners.
Still, for this retro bakery, Rubin can't help but add some future elements to it. He manually coded the POS system to accept payments and record transactions without the intervention of a cashier. In addition, Rubin also set up a private meeting room in the backyard of the bakery, and built a homemade magnetic lock system.
Rubin’s time at this meeting was not short, as he often invited friends to taste the various pastries and was also discussing a question: What to do next?
Leading mobile computing Rubin's career has been at the forefront of the mobile computing revolution. In 1992, he worked for General Magic (a company that was spun off from Apple) and led the development of the first wireless PDA product "Motorola Envoy." In 1999, Rubin left Microsoft and founded his own company Danger, which later invented the T-Mobile Sidekick.
In 2003, Rubin co-founded Android with others. At the time, Android was just an inconspicuous open source system that was sold to Google in 2005. At the time, the smartphone industry was still in a chaotic situation, and smartphones were not compatible.
After giving up the Android operating system, Rubin provided a common language and a complete set of tools for smartphone manufacturers, thus promoting the popularity of smartphones. Today, Android is one of the fastest-growing consumer electronics technologies in history, running 25,000 different products, including mobile phones, tablets, smart watches, smart TVs and smart bracelets.
Rubin has led the development of smartphones from concept to popularity. Today's smartphones are not very attractive to Rubin. Because it has been solved as an engineering problem. It is true that an entrepreneur may continue to introduce new applications. But for a person who regards engineering as an art, this is equivalent to adding a few more pens to the already painted oil paintings. Rubin hopes to recreate one.
AI will become the next computing platform Rubin has always believed that human beings are always on the cusp of the new technology era. Just as MS-DOS gives way to Macintosh and Windows, then give way to the web and smartphone. Rubin believes that the next great computing platform will be artificial intelligence (AI), a process that can take decades.
Companies such as Google, Facebook, and Microsoft have spent billions of dollars developing neural networks that understand human language or recognize faces. In the next 10 years, artificial intelligence will grow stronger and be able to perform complex tasks that are still unimaginable today.
Rubin believes that artificial intelligence will soon develop into a cloud service that powers thousands of devices and machines. This is like having some software installed on every device today. It will be difficult to buy a device without pre-installed artificial intelligence technology in the future.
As for what the future will become, it is not yet possible to predict accurately, but I can imagine the difference between ordinary cars and self-driving cars. We can also borrow from today's devices to imagine the future: today's TVs can translate programs into any language in real time, a monitoring system that automatically recognizes family and thieves, and the oven knows when food is grilled it is good.
In 2013, Rubin made a long talk with Larry Page, and the two agreed that it was time for change. In March of that year, Rubin was no longer responsible for the Android business, and later turned to Google's newly established robotics department. But Rubin quickly realized that the goal of the Google Robotics department required a 10-year foundation study.
Rubin can't wait that long. Rubin's friend and consultant Marc Andreessen said: "Rubin can't tolerate the state of the world. He often thinks about the social situation after 5, 10, or even 15 years, and then looks back at today's world. And found that there is still a lot to do."
Rubin hopes to use the sensing machine to provide data to the neural network to fill the world.
Create a new company Playground Global
In 2014, Rubin left from Google. A few months later, Rubin launched the new company Playground Global. This is not only a new company, but also a new type of company. It is like an incubator and a consulting company, but in fact neither.
Playground Global invests in hardware startups, but not just providing funding and advice. Rubin also offers them an All-Star engineering team consisting of Google, General Magic and Apple's former engineers.
Playground Global's mission is not limited to building a product or a startup. Rubin wants it to be a factory that produces standard building modules for the future of artificial intelligence. He wants to open this platform of software and hardware tools to everyone to create smart devices.
If successful, Playground Global's impact on future artificial intelligence machines is equivalent to Android's impact on smartphones, providing a technology infrastructure for thousands of products, giving entrepreneurs the ability to build smart drones, smart appliances and full-featured robots. Ability.
According to Rubin, the fundamental idea is to create a "creative amplifier," a system that quickly transforms concepts into products. For a person like Rubin who is dissatisfied with the status quo of the world, this is a suitable goal, hoping to accelerate the realization of future goals.
Rubin has cultivated this curiosity since 1978. At the time, he was a high school student and used a computer to remotely control Kenner's electronic toy R2-D2. The Danger created by Rubin in 1999 and later Android fully demonstrate its love for robots. The name "Danger" comes from the space fantasy TV series "Lost Space", which was popular in the United States in the 1960s. In the play, a robot often issued a "Danger!" warning in the play.
Information from the AI ​​era comes from the physical world back to Rubin's new project artificial intelligence. At present, efforts on artificial intelligence are mainly focused on building large-scale neural networks, a task involving the collection of massive amounts of data. Currently, most of this data comes from the Internet.
But Rubin believes that in the future world of artificial intelligence, this information should come from the physical world. To achieve this, you need to create a large number of devices to gather information from your surroundings: text, images, sounds, geography, weather, and other data.
Rubin hopes to use these data collection machines to fill the world to train these large-scale neural networks. This will create a benign feedback loop: new machines will make artificial intelligence smarter, and more advanced artificial intelligence technology will allow these data collection machines to better collect information.
Today, it's easier than ever to build such a smart device. Sensors and CPUs are getting cheaper and cheaper, and Chinese manufacturers are eager to OEM for startups. But "it's easier than ever" is also accompanied by "difficult challenges" because there is a huge difference between building a prototype and a formal commercial product.
This is true even for the simplest hardware products. If we want to produce a car camera, the first thing you need is the SOC (system-on-chip, including CPU and peripherals), you may also need LCD, camera module (including imaging sensor), battery management battery connection, GPS connected to the antenna Chip and wireless interface (requires another antenna) and so on.
These are just electronic systems, and we have to consider a lot of issues in terms of design and software. For example, how to make products as small as possible, how to be more energy efficient, and how to get more reliable parts at a lower price. In addition, it may take 1-2 years for the product to go on the market, and today's parts may be outdated. Therefore, we also want to predict how the parts will be after one year and design the products accordingly. If the prediction is wrong, this car camera may have been eliminated without going public.
For each of these issues, every hardware entrepreneur must consider it. Fortunately, Rubin has had such an experience before. Smartphone makers are facing the same problem before opening Android. However, after the opening of Android, mobile phone manufacturers were freed from it and launched more models.
This is what Rubin wants Playground Global to do: provide all the basic hardware and software components so that entrepreneurs can focus on developing interesting devices. If you want to develop a drone that requires the best microphone array, then the excellent engineers at Playground Global will provide it to you. Rubin said: "This will be a modular hardware. After a few years, if you find us with ideas, we will provide you with these modules."
Rubin said that in the short term, the platform will only be open to startups investing in Playground Global. The goal is to help these startups build products faster than competitors. But in the end, Rubin hopes to open the platform to everyone, just as Android is available to all mobile phone manufacturers. By then, everyone can use Playground Global's tools to create a new generation of smart hardware.
If you do this, Rubin's company will benefit in two ways: First, when entrepreneurs are ready to set up their own company, they may want to work with Playground Global to win investment opportunities. More importantly, they will also get the technology of Playground Global.
This is also Playground Global's dream: to create a common architecture for thousands of devices, just as Windows affects PCs and Android. Jeff Brody, an investment expert at Redpoint Ventures, said: "These standardized frameworks will open the next generation of interesting smart devices."
The order of development of the Internet of Things has been reversed. Today, Rubin’s idea may not sound strange. In the past few years, technology companies have realized the arrival of the Internet of Things era. Google, Apple and Samsung are all building their own ecosystems, allowing device vendors to develop products around their platforms.
But Rubin believes that these companies have reversed the order. Ecosystems are used to support popular equipment, and equipment should be available before the ecosystem. Therefore, Playground Global first provides technology for startups to develop new devices, and the ecosystem will then emerge. Rubin said: "We invest in providing modules and technology that are planting for the next ecosystem."
For the future world that Rubin imagined, he said that waiting for this moment is like blocking the road. He said: "When I have an intuition about the future world, I try to find out why today has not been realized."
This is why Rubin created Playground Global. Currently, several startups have settled in the Playground Global headquarters. Bruce Leak, a Rubin friend and co-founder of Playground Global, said: A basic principle is that there is no need to obtain any permission here. Playground Global assigns a desk and meeting room to each company, and people can drag it at will. The network cable hangs from the ceiling and can be used at will. If you need 3D printed parts, you can use the two labs arranged by Playground Global.
In addition to Rick, Robin has two other co-founders, Peter Barrett and Matt Hershenson. The former is Rubin's colleague at General Magic, the latter is the co-founder of Danger. Last year, Playground Global successfully raised $300 million, including Google, HP, Foxconn, Red Dot Ventures and Tencent.
Not satisfied with acting as a planner Rubin is never content to act as a planner. Therefore, in addition to Playground Global, Rubin himself is also building some products. Since Playground Global is developing software and hardware to serve the startups it invests in, why can't Rubin use these technologies to develop his own products?
Rubin refused to disclose his plan. There have been reports that Rubin is developing a new Android phone, but Rubin declined to comment. He said that he is developing a car camera and has other ideas, but it is not yet revealed.
Rubin’s vision of the future world may be different from others. The philosopher Nick Bostrom believes that the continued development of artificial intelligence will be a potential threat, and this view is also obtained by Elon Musk, Stephen Hawking and Bill. · The identity of Bill Gates. But Rubin doesn't think so. He said: "Overall, I believe that technology is being used more in good faith."
For Rubin, the more immediate challenge now comes from the competition of some large technology companies. Microsoft has set up a platform to collect and analyze sensor data, which is reported to collect one trillion pieces of data per week. In addition, Google also owns the open source artificial intelligence software engine TensorFlow, Facebook is also working on artificial intelligence, and Musk also created OpenAI, a non-profit organization focused on artificial intelligence.