Artificial intelligence and job threat .. Deep analysis and surprising results

With the development and adoption of artificial intelligence techniques in many fields, many have begun to wonder about the impact of this development on their future employment opportunities.

No one is yet able to predict the impact of artificial intelligence on the future of work, but there are two views on the subject: the first says it is destructive and will cause disruption in the labor market and predicted future robots to take their functions from humans. A second view is that automation can create more jobs by offering new tools to entrepreneurs, and it may also create new business lines that we can not now envision. This view is corroborated by a recent study by Redwood Software for Digital Transformation Projects, Participants believe that 60% of companies will be automated over the next five years.

Gartner Research predicts that "by 2020, artificial intelligence will produce more jobs." "I look at our company today to see that two-thirds of its jobs today have not been around for a few years," said Dennis Mortensen, chief executive and founder of x.ai, a manufacturer of artificial intelligence assistants.

In addition to creating new jobs, artificial intelligence will help humans do their jobs much better, said Paul Dowherty, head of technology and innovation at Accenture at the World Economic Forum in Davos. He summarized the future of automation in one phrase: "Man + Machine equals super powers. "

So how will artificial intelligence increase jobs? Create new jobs and improve existing ones
For many reasons the optimistic view is likely to be the most realistic, but the ability of artificial intelligence to transform the future of work is far from being settled.

Dugherty and Jim Wilson, executive director of information technology and business research at Accenture Research, co-authored the book "Human + Machine: Reimagining Work in the Age of AI," which argued: "The technical and managerial functions required by artificial intelligence in the future Currently, there are two categories: Instructors and Charges. Instructors will teach how artificial intelligence systems implement and mimic human behavior, and will create connections between human machines and supervisors by explaining how complex algorithms work. "

Chatbots have recently emerged as a new channel of communication for brands and consumers. No one is afraid that they often provide inappropriate responses. Here, the class of trainers who teach artificial intelligence algorithms shows how to simulate human behavior by developing natural language processors and translation systems to reduce The percentage of errors is no doubt that understanding the language is one of the most important factors required in interactive Chatbots, but there is another factor that is no less important. It is a factor of empathy. This is evidenced by the fact that many startups are keen to have an emotional intelligence factor in the Continue based on interactive chat robots.

"The chat robots of our company that understands human emotions are based on human trainers," said Eugenia Kuida, co-founder of Repika. "In the future, I think the field of knowledge of human behavior and psychology will be the most exciting field of knowledge." And they want to achieve their goals, and this will not happen without a certain amount of sympathy. "

In addition, companies such as Facebook and Google use humans to manage and control content. Facebook currently has about 7,500 people. Google's parent company, Google Alpha, recently said it plans to involve 10,000 people to manage and monitor content on YouTube.

For example, the evolution of artificial intelligence systems makes them more ambiguous, and this is a concern for many senior executives because of the complex nature of the Blackbox-like learning algorithms, especially when the system is likely to take Acts that conflict with axioms, so they will need free English explanations of the technical complexities of an ordinary human being to reduce their fears.

Legislation is also an important factor influencing the future of automated learning algorithms. The EU has issued a new general data protection regulation called "Right to Interpretation" which will come into effect this year. This means that consumers can question and oppose any decision made using algorithms and have This is where the interpreters come to perform autopsy processes when the machines commit errors, they will also diagnose errors and help take steps to avoid similar errors in the future.

Artificial intelligence supports workers, businesses and industries
Instead of replacing workers with artificial intelligence, it can be used as a tool to help employees work better. For example, a call center employee can use it to get instant, comprehensive information about what a caller needs and do his work faster and better. This applies to most companies and industries as well. In another example of life sciences, Accenture Research uses deep learning and neural networks To help companies deliver solutions to markets faster. In addition to already existing companies, artificial intelligence can create companies working in new areas, such as companies that offer digital solutions for elder care, artificial intelligence agriculture and monitoring of sales calls Built on the basis of artificial intelligence.

"There is a shortage of 150,000 truck drivers in the United States right now," said Dougherty. "We need automation to improve drivers' productivity and lifestyle to attract more people to the profession."

The changes we must make at the moment
It is likely that it will take ten years for artificial intelligence technology to become the norm, because so far there are few companies that have begun to adopt artificial intelligence techniques by taking measures to train their workers. Accenture's research indicates that business leaders do not think their workers are ready for intelligence Now artificial, but only 3% of them are considering investing to train their workers on these techniques.

On the sidelines of the Accenture conference at the Davos World Forum, Fei Fei Lee, a Stanford University professor and director of the Artificial Intelligence Laboratory, suggested that artificial intelligence can be used to retrain workers.

"I think there is an impressive possibility that automated learning can help us learn more efficiently and rehabilitate workers in more effective ways, and personally I would like to see more investment and reflection on this aspect," she said.

Another thing that needs to be addressed in 2018 is the lack of real diversity in companies producing artificial intelligence technology. As I pointed out, this lack of diversity is a bias in itself. "

Recent research by the Massachusetts Institute of Technology has confirmed this point. "I found evidence that facial recognition systems recognize white faces better than black faces," said researcher Joey Bolamwini. In particular, the study found that if the image of a white man is, the systems estimate more correctly where the proportion is 99% most of the time, but for black women it ranges from 20% to 34%. These biases have implications for the use of facial recognition techniques Especially in the areas of law enforcement, advertising and employment.

In the end, this report shows what artificial intelligence can not be considered as a strange force that presents many disorders on its own, but is in fact a human invention that reflects the flaws of its inventiveness as well as its human side.

"The impact of artificial intelligence on jobs is entirely within our control," said Cathy Bessant, head of operations and technology at Bank of America, at the Davos Forum. "Artificial intelligence will not affect the future of human functions, but is a way of developing future jobs.