Location: Home >Posts> Is it the best solution to equip human workers with robot brains?

Is it the best solution to equip human workers with robot brains?

AsiaIndustrial NetNews: With a soft whirring sound, it crossed the warehouse floor, raising or lowering its two arms on a scissor lift, ready for its next mission, according to the BBC. With cameras mounted on joints on each arm, the left hand pulls the carton out of the shelf, and the right hand pulls a bottle out of it.like many newrobotThat way, this robot is from Japan. Hitachi showed it in 2015 and hopes to go on sale in 2020.

Is it the best solution to equip human workers with robot brains?

Figure: A robot designed by Hitachi, Japan

It’s not just a robot that can remove bottles from a shelf, it’s a robot that can perform such seemingly simple tasks with the speed and precision of previous workers. Right now, humans and such Robots are working together in warehouses. In the future, they may replace all warehouse workers. Kiva robots can be found everywhere in Amazon warehouses, not only taking things from shelves, but also carrying shelves to humans so they can pick items. In this way, the Kiva robot increases work efficiency by a factor of 4.

Figure: 45,000 Kiva robots have been deployed in Amazon warehouses

Robots are also working alongside humans in many factories. Robots have been used in factories since 1961. That’s when General Motors deployed its first robot, the Unimate, a one-armed robot that could perform jobs such as welding. But it was only in recent years that robots began to be strictly segregated from human workers, partly to protect humans and partly to prevent human workers from interfering with robot work, whose working conditions had to be strictly controlled. With modern robots, these controls are no longer necessary.

Pictured: General Motors’ Unimate robot

Launched by Rethink RoboticsindustryTake the robot Baxter, for example, which usually avoids colliding with people. Even if someone bumps into it, it won’t fall to the ground. Cartoon-like eyes can give human colleagues hints about the direction in which they are about to move. Historically, Industrial robots required professional programming, but Baxter can learn from human workers how to perform new tasks.

Pictured: Baxter, a robot from Rethink Robotics

1 2 Next page > page

The global number of robots is increasing rapidly, with Industrial Robot sales increasing by 13% in one year, which means that the “birth rate” of robots doubles every five years. The long-standing trend is that cheap labor in emerging markets has always attracted “offshoring manufacturing.” Today, robots are experiencing a “return trend”, re-attracting manufacturing to established manufacturing centers.

These robots can do a growing number of jobs, such as picking lettuce, bartending, and transporting in hospitals. But they still fell short of our expectations. A year after Unimate was launched (1962), Rosie, the robot that appeared in the American cartoon “Modern Family (The Jetsons)”, could do almost all housework. However, it seems that the realization of this goal is still far away.

Figure: Rosie the robot in “Modern Family”

These advances are largely thanks to improvements in Robotic hardware, including better and cheaper sensors, especially improvements in robotic eyes, fingertip haptics, and balance. In addition, there has been great progress in software, and the brains of robots are getting smarter and smarter. And over time, machine thinking is evolving into a new field that had high hopes and early disappointments.

Attempts to invent artificial intelligence (AI) date back to 1956, when at a summer symposium at Dartmouth University, AI pioneers argued that machines should be able to use language, form abstract thoughts and concepts, and solve human problems. problems, and self-evolution. They also believe that machines with human-like IQs should be available within 20 years. However, they now think it will take another 20 years to achieve this goal.

The futurist and philosopher Nick Bostrom once questioned this: Self-planning = superintelligence? “For prophets, 20 years is the sweet spot for predicting a sea change,” he writes. The closer you get to this point, the more you’d expect to see a prototype, and the farther away it’s less likely to attract attention. In the past few years, advances in AI technology have begun to accelerate. Especially those that focus on one thing, like Go, filtering out spam emails, or AI algorithms that recognize faces in photos.

Pictured: Google AlphaGo beats South Korean Go champion Lee Sedol

Processors are getting faster, datasets are getting bigger, and programmers are getting better at writing algorithms that can learn how to improve themselves. Yet this capacity for self-improvement worries many thinkers. What if we could create artificial intelligence that solves any problem like humans can? Will it quickly become super smart? How do we make sure it’s under control?

This is at least not a looming concern now that super AI equivalent to human intelligence may be another 20 years away. But AI in the narrow sense is already starting to change the economy. Over the years, these algorithms have taken over white-collar jobs in many fields, such as bookkeepers and customer service. More and more decent jobs are also under threat. IBM’s artificial intelligence Watson made headlines by beating a human champion at a mind game and has now surpassed human doctors in diagnosing lung cancer. The software was even better than an experienced human lawyer at predicting which cases were likely to be won. Robo-advisors provide better investment advice.

Pictured: IBM’s AI beats human champions at mind games

Algorithms regularly make news stories in financial markets and sports events. Lucky for me, they don’t seem to be able to write op-eds on tech and economics the way humans do. Some economists estimate that robots and artificial intelligence can explain strange economic trends. Erik Brynjolfsson and Andrew McAfee, MIT professors and co-authors of the best-selling book “Race against the Machine,” argue that there is a “huge decoupling between work and productivity” ”, that is, how efficiently the economy has low inputs, such as people and capital, to turn them into something useful.

Historically, higher productivity has meant more jobs and higher wages. But Brynjolfsson and McAfee argue that this model will no longer apply in the United States, because in the 21st century, while U.S. productivity has greatly increased, jobs and wages have failed to keep pace. Some economists worry that we are experiencing “secular stagnation,” where there is not enough demand to stimulate economic growth, even with low or zero interest rates.

New technology eliminating jobs is nothing new. 200 years ago, the Luddites were desperate to disrupt technology. Today, Luddites have become the butt of a joke, because technology ultimately creates more jobs to make up for those it replaces. And provide better jobs, at least different ones. The current consequences of robotics are unclear, and it is likely that some human jobs will be threatened.

Pictured: Jennifer Unit is a voice-guided computer program that guides workers on how to better perform tasks

That’s because the robot’s brain seems to be progressing faster than the body. Robots can help take off and land planes and trade stocks on Wall Street, but they can’t clean bathrooms, says Martin Ford, author of Rise of the Robots. Perhaps in the future, we should not expect robots like Rosie in warehouses, but another smart device, the Jennifer Unit.

The Jennifer Unit is a computer headset that guides human workers on how to perform tasks better, with more attention to detail. If you need to pick 19 identical items, it will tell you to group 5, 5, 5, 4, which is less error-prone. Since robots may outperform humans in thinking, and humans beat robots when it comes to picking things from shelves, why not use robotic brains to control human bodies? It might not be the perfect career choice, but you can’t deny that it’s logical.

The Links:   JANCD-YIF01-1E 3HAC031670-001 IGBT