People and robots: 4 points of contact

People are no longer worried that robots will kill them in the future. However, what will be the life of humans and robots together? At the Next: Economy conference, they named 4 areas in which people will have to “cooperate” with robots in the near future.

Although the idea of ​​the coexistence of humans and robots may seem dubious, we are unlikely to be able to avoid this. This idea was key to the topic “Minds and Machines,” which opened the Next: Economy conference in San Francisco. The topic of automation and how it will affect (or will not affect) our lives was highlighted in a number of conversations with relevant experts.

Here are 4 main points of contact between humans and robots in the future.

Cars

The first conversation touched on the topic of autonomous, robotic cars. Interviewed by Stephen Levy of Backchannel and Sebastian Trun, director and co-founder of Udacity. Troon, who studied at Stanford University, was personally invited to work at Google by Larry Page to work on robotic cars. Trun remembered that when Page shared the idea of ​​creating robotic cars with him, he replied that this was impossible. Page asked why. It was after this that Trun asked himself this question himself and came to the conclusion that, if no one had done this before, this does not mean that it is not possible to create a fully robotic, self-propelled car.

Levy asked Trun a question about how the introduction of robotic cars into the system will affect the economy. Will taxi drivers lose their jobs? To this Trun replied that robotic cars will reduce the number of accidents and deaths on the roads, reduce gasoline consumption and allow people to live much further from work.

Reflecting on how robotic cars will reduce the number of fatal accidents, Trun touched on the subject of machine learning. When a driver makes a mistake on the road, he learns and, most likely, he (and only he) will not allow it anymore.

“In the world of robots, everything is different: if a robotic car makes a mistake, it is immediately entered into the database, and no other car will ever make it again,” said Trun. Since machines learn much faster than people, the collected data can be used in other areas.

Finally, Levy raised the theme that many people do not trust cars. Trun retorted that, although it’s often difficult for people to admit to themselves whether they trust cars or not, they depend on them, whether they are smartphones, applications, or bank accounts.

Personal assistants

In the next section, Levy talked with Adam Cheyer of Viv (the company that created Siri) and Alexander Lebrun of M (personal assistant recently acquired by Facebook).

Cheyer introduced the idea of ​​a post-mobile world in which personal assistants will be transferred to the “cloud”, and people will be able to make complex queries, such as “On the way to his brother, you need to choose a good wine that is suitable for lasagna.” Viv will go through all sites and services and determine how to quickly and conveniently get to a given point, which wine shops should go, which wine is best suited to a particular Lasagna recipe, etc.

Viv is “a new way to work with sites and services on the Web around the world,” said Cheyer.

M, in turn, will be a personal assistant built into Facebook Messenger. With its help it will be possible to order flowers, tables in restaurants and hotel rooms, send requests to the telephone company, etc.

Levy asked Cheyer and Lebrun the obvious question of whether technology would take jobs from people. Lebrun said that, on the contrary, M will need a trainer and many domains, so in a sense, additional jobs will be created.

Emotions

Then Levy called on the carpet John Markov of The New York Times and Jerry Kaplan of Stanford University, the author of the documentary Humans Need Not Apply .

Markov spoke about his years at Stanford, where there were two camps holding different views on artificial intelligence - the former worked on completely autonomous systems, the latter used artificial intelligence to enhance people's abilities. Both camps did not particularly communicate. Markov monitored the process in both the first and second camps.

“They had a choice whether to create people of the future or not,” he said.

When it came to boss robots and their relationships with people, Levy recalled the 2013 film “She,” in which a man falls in love with the operating system.

Cheyer said that when he watched the film, he constantly tried to analyze and predict that the OS would respond to the engineer. However, when “she” began to act too emotionally, he lowered his hands.

“In reality, we are still very far from the moment when the machines have emotions,” said Cheyer.

The participants talked about the tendency of people to endow emotions and humanize everything around. Markov said that in fact, the Turing Test is a test for the gullibility of the human race.

“The requirements are incredibly low. We are ready to humanize anything, ”he said.

Then it came to the sensational report of the University of Oxford that 47% of all professions in the United States could be automated.

Both Kaplan and Markov retorted that many professions require human participation.

“Nobody wants to come to the undertaker, who will say that he regrets the loss of a robot in his voice,” Kaplan said. The point is completely different: we need to automate not professions in order to replace people, but separate tasks and processes, the participation of people in which is not required.

Analysis

The cause of tension between humans and robots used to be obvious, says Christian Hammond of Narrative Science. It was expected that they would just kill us all. Now, however, the problem is even more serious - robots are going to take away our jobs.

Hammond talked about how Narrative Science and the Quill service developed by the company made a splash by collecting data on the company's quarterly reports and generating a full-fledged journalistic article for Forbes. It was expected that the profession of a journalist signed a death sentence.

“We can analyze what and how they do, by collecting the necessary data, understand what we need to say, and then put it all in writing,” Hammond said.

However, flirting with the data did not lead to the creation of a legion of robotic journalists, but to the automation of the work of the analysts themselves. Hammond explained it simply: so much data is being generated in the world that no analysts are enough to process it.

In addition, we are living in an era of personalization, and it may make sense to use Quill to analyze small companies' credit card statements and generate suggestions on how to spend money more efficiently and not write articles instead of people.

You can also use Quill to generate comments for financial portfolios or reports on the operation of touch IoT devices.

“Data analysis was the most attractive profession of the 21st century. Now, this is just the next candidate for automation, ”said Hammond.



Translation prepared by: greebn9k (Sergey Gribnyak), silmarilion (Andrey Khakharev)