Rethink Robotics founder's prediction of future technology
Author: Co-founder of Rethink Robotics, Chairman and CTO Rodney Brooks
For all new technologies, there will be predictions like "what benefits will this technology bring to humanity" or "how bad the technology will be." One thing I found in common is that people always overestimate the short-term benefits of a technology, but underestimate its long-term effects. This has been pointed out in my previous post, the Seventh Deadly Sins of Predicting the Future of AI*.
I think there are a lot of speculations about some of the predictions of artificial intelligence (AI) and machine learning (ML). In the past few months, I have also poured a bit of cold water on it, hoping that everyone will look at these technologies more rationally. I am not a technical pessimist. Instead, I consider myself a technical realist.
I feel that it is easy to know. It's hard to turn ideas into reality. It is even more difficult to make an idea that can be applied on a large scale. However, after evaluation, based on the possibility of success, we still have the possibility to classify these technologies and technology applications from “relatively easy†to “very difficultâ€.
There will be a lot of predictions about the coming year at the beginning of the year. I also took this opportunity to express my predictions, not only for the coming year, but for the next thirty-two years. The farthest date I predicted was January 1, 2050. This means that I will predict the technological development of the first half of the 21st century.
Regarding the date, I have three different prediction methods:
NIML (Not In My Lifetime): I can't see it, after January 1, 2050.
NET (Not Early Than) date: no earlier than this date
BY date: before that date
Sometimes, I use both NET and BY dates for a certain forecast, and I believe this prediction will become a reality.
My prediction rules
I will strive to make a very accurate description of these predictions and time. In fact, it is almost impossible to precisely define the development I have predicted. But I will try.
My recent experience has made me realize how people will adhere to preconceived technical concepts in the face of challenges. I said this on Twitter:
When humans land on the moon for the next time, they will rely on the technology of many artificial intelligence and machine learning systems.
Last time we went to the moon, there was no artificial intelligence or machine learning.
I want to express that although artificial intelligence and machine learning are very powerful and useful today, it does not mean that this is the only way. It also shows that everything in the world does not change in some magical way.
That's why I tried to make a very specific description of my predictions. Moreover, I will refute many people who will claim that my predicted development "will not happen before a certain year" has already happened. I predict that someone must do this!
What is easy? What is difficult?
It's relatively easy to make electric cars and reusable rockets; flying cars, or the Hyperloop super high-speed rail system (or similar underground traffic network) is a difficult task.
Where is the difference?
The car appeared a century ago and is now in mass production. If you want to upgrade from a petrol car to an electric car, we don't have to create too many things, and we don't have to spend a lot of time to promote the application on a large scale.
Most of the car's components, including wipers, brakes, wheels, tires, steering systems, windows, car seats, chassis, etc., we have more than 100 years of engineering knowledge and production experience. Moreover, we have more than 20 years of experience in mass production of digitally driven trains.
Reusable rockets look revolutionary, but they also rely on existing technology and experience. The main components and functions of all liquid fuel rockets are now similar to the V-2 rockets designed by Wernher von Braun during World War II. High-flow turbo pumps (580 hp) using liquid fuel, fuel, cooling components for the engine, and liquid oxygen carried are all 75 years ago. And also achieved mass production. People built 5,200 V-2 rockets in just two years.
Since then, there have been more than 20 different liquid fuel rockets around the world, some of which have been in use for more than 50 years. With different models and configurations, there are no fewer than a few hundred members of this family.
We have invested heavily in the development of rocket technology, and have invented many available technologies, intellectual property, and accumulated a wealth of experience.
Of course, this is not to say that scaling an electric car or propelling a reusable rocket is a non-creative job or a breeze. These advances are important, but they are based on the results of their predecessors and are therefore more likely to succeed.
However, for new ideas, it is more difficult to predict with certainty at which point in time when these technologies will be applied.
The concept of Hyperloop's super high-speed rail has attracted a number of emerging companies and capital, but there is no practical proof of this concept, let alone large-scale applications. Therefore, we have studied how to develop ultra-stable vacuum pipes of hundreds of miles, as well as capsule trains that are driven by external air pressure to accelerate at hundreds of miles per hour and ensure passenger safety. There is still a long way to go. Going.
For this concept, although there may be some major breakthroughs in the next 32 years, I believe that there will be no commercial ultra-high-speed passenger system in this time frame.
About new technology applications
New technologies take longer than we expected from development to application.
Initially, the Internet used 32-bit addressing, using the IPv4 Internet protocol, providing an IP address of approximately 4 billion. But in the early 1990s, people realized that in addition to personal devices, devices such as electricity meters, industrial sensors, flow sensors, controls, televisions, light switches, etc. would be added to the network, and the Internet address space would be exhausted.
Subsequently, IPv6 was proposed in 1996 to increase the address space from 32 bits to 128 bits to accommodate more network devices. Since 1996, our target date for IPv6 replacement for IPv4 has been changing. For example, in 2010, this target date was 2012; but by 2014, 99% of network traffic still uses the IPv4 protocol. In 2017, devices using IPv6 addresses increased from 2% to over 20%. But there is still a long way to go before the full adoption of IPv6.
At the technical level, nothing can stop the application of IPv6. And on the contrary, if we want to connect a larger number of devices to the Internet, we need to do a lot of smart innovations and workarounds to work with IPv4 instead of IPv6.
About "Time is always longer than expected"
SpaceX Corporation announced in 2011 that it will launch the development of the "Falcon Heavy" rocket and is expected to launch for the first time in 2013. However, the rocket was erected vertically at the Kennedy Space Center's 39A launch platform in December 2017, followed by a series of tests to prepare for the first flight in January 2018. So far, the first flight time has been postponed from two years after development to seven years.
The time from development to application of these new technologies is always longer than we expected.
Forecast for autonomous vehicles
The first three items in the table below are about flying cars. I can say with certainty that a flying car that is going to be used will require a lot of autopilot technology, so it is appropriate to attribute it to the autonomous driving category. The flying car I defined must be able to fly anywhere the current car can reach, otherwise it will not be a car. Moreover, I mean that the driver of a flying car does not need a pilot license, and may only need a few hours of special training to wear a normal casual wear/flying 100 miles. Of course, most of the journey is in the air. In addition, this does not require prior arrangement or special application, just like we use a smartphone for navigation driving today.
Autonomous vehicles don't just refer to ordinary cars without drivers. They are used in different ways, and the way they integrate into this society will be different. I am very clear about this.
This is like a car that is not simple without a horse. Driving a car requires a new road infrastructure, ownership patterns and usage patterns are different, and fueling and maintenance are more different.
I don't think that autonomous cars will completely replace human drivers. Instead, our cities will make corresponding changes, such as opening up dedicated self-driving car lanes, building geofences to locate autopilots and drivers, making changes to pick-up points, parking rules, and more. In short, our city will have a variety of subtle changes.
Prediction of robotics, artificial intelligence and machine learning
Some of the predictions in the table below include public perceptions of artificial intelligence, technical ideas, and application deployment.
The general intelligence that can suddenly achieve the level of human (or chimpanzee) does not exist. It will be a long and long gradual process, and it will be improved little by little.
Predictions about space travel
My predictions about space travel below may not be as optimistic as I hoped, but more realistic.
About Rodney Brooks
Born in Australia, Rodney Brooks received an undergraduate degree in mathematics from Flinders University in South Australia and a Ph.D. in computer science from Stanford University. Between 1984 and 2010, Rodney Brooks taught at the Massachusetts Institute of Technology and was a renowned robotic professor. He founded the Massachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory and has served as Director until 2007.
In 1990, he co-founded iRobot (Nasdaq: IRBT), where he served as chief technology officer, chairman and board member until 2011.
He was elected to the National Academy of Engineering and is a member of several leading industry bodies, including the American Academy of Arts and Sciences, the American Computer Society, the American Artificial Intelligence Association, the Institute of Electrical and Electronics Engineers, and the American Association of Advanced Science.
Rodney Brooks is also committed to promoting robotics and artificial intelligence, and is active in all major events around the world.
About Rethink Robotics
Through its intelligent and collaborative robots, Rethink Robotics can continue to innovate manufacturing methods by completing the work that 90% of the traditional automation solutions can't. Supported by the Intera software platform, Baxter and Sawyer robots can adapt to the variability of the real world, flexibly and quickly switch between different application scenarios, and complete tasks like people. Manufacturers of all sizes and industry types can enjoy rapid deployment, easy-to-use and smart automation solutions that increase production flexibility, reduce costs and drive innovation.
6Gang Switch And Multi Function Socket
6Gang Switch And Multi Function Socket,6Gang Switch,Multi Function Socket,Multi Function Socket Wrench
ZHEJIANG HUAYAN ELECTRIC CO.,LTD , https://www.huayanelectric.com