The manner in which people talk about how jobs and work are changing because of Artificial Intelligence (AI) and automation does not match up with the reality, based on a number of speakers at MIT’s latest Artificial Intelligence and the Work of Future Congress.
In the panel on the misconceptions of Artificial Intelligence technology, a number of speakers talked about how Artificial Intelligence cannot do everything do, and just how humans have to be an important part of the new technologies being developed.
Julie Shah, Associate Professor in MIT’s Department of Astronautics and Aeronautics, said a myth was that people are able to simply set up automation in factors depending on the data we’ve. She has invested time on factory floors, observing associates practicing how you can set up a new model.
“We do not understand how to set a [manufacturing] unit is probably the most optimum way,” she said, and that is why “lights out” factories do not show improvements. Rather, it was humans constantly iterating on the procedure based on changing conditions that have been more successful. Moving forward, methods have to fully understand us, while we have to be aware of how systems are behaving.
Shah said we bring so much knowledge, background, to a choice that it’s extremely hard to codify or describe. She noted that with several of today’s versions, explaining the decision making is often difficult. Instead, Shah suggested using “domain experts” to guide a machine’s inference procedure, helping to figure out implicit ideas and prioritization.
Jobcase CEO Fred Goff stated we’d a chance to harness Artificial Intelligence “for the real empowerment of people,” rather than simply using it to replace tasks or workers.
Goff claimed that technologies have been displacing jobs for the last half-century and that the huge problem was wage stagnation and under-employment. We should do realize that Artificial Intelligence and automation are able to do tasks, not jobs, therefore they cannot replace everything a human can do. We need to consider “machines and humans, not machines Or humans,” Goff said.
Both Shah and Goff agreed that Artificial intelligence should not determine results, but rather can certainly be best used as an option of tools for decision-making.
“Retraining and reskilling do not mean humans have to sit down in front of computers and code,” Goff said. He noted that workers talk regarding retraining coal miners to code and related ideas. Rather, he said we need to have broad post-secondary training for humans but that we need to realize that “not everybody has to go to college.” There’s a huge need for people in the trades, like plumbers and welding, he stated, and thought about if there was a larger opportunity for “micro-certification.”
James McGlennon, Chief Information Officer of Liberty Mutual Insurance was questioned by MIT’s David Autor regarding the way, how jobs are changing due to Artificial Intelligence. McGlennon stated he was noticing tons of change, but it was due to “agile transformation” and “business agility” rather than Artificial Intelligence.
He stated that Liberty Mutual is seeing more need for higher-skilled jobs, which specialization, where technology is a crucial driver, is “coming into vogue,” while in performs where technology is not the key feature. As a result, everybody has to understand technology a lot better. He says that Liberty Mutual associates throughout the board know that to be able to stay relevant, lots of people should reskill themselves, together with the company offering resources and training.
He agreed with an earlier comment which abilities are starting to be much more critical compared to credentials and that skills are definitely more important in even more jobs. “What will differentiate winners and leaders,” he said, is going to be their “ability to recognize and interrelate with humans at every level.”
Scott Prevost, Vice President of Engineering for Sensei at Adobe was adamant that Artificial Intelligence today “is empowering the worker,” amplifying the innovative person’s experience by automating the points people have to do but do not wish to do. 74% of Adobe’s customers stated they invested half their time carrying out repeated, non-creative tasks.
Prevost was looking ahead to an innovative assistant and a marketing assistant to enable you to move through your entire workflow. He stated that marketers and “creatives” won’t go away, though their roles might shift. A creative will get much more of an art director instead of doing fine-grain production. As an outcome, the focus is going to shift to innovative problem solving, being innovative, and collaborating effectively.
This can make designers try out more ideas per client, allowing them to try out the most innovative ideas, while simultaneously reducing the bar for less-skilled employees.
Zeynep Ton, professor at the MIT Sloan School of Management, observed that technologies have constantly had a deep impact on the retail industry, noting that probably the biggest change in the past few years has been e-commerce. However, consumer experience and employee knowledge have not changed a lot. She noted that retail is probably the largest employer in the land, but is recognized for bad jobs, wages that are low, unpredictable schedules, and “people employed as widgets.” She stated we have a chance to redesign work, but and so much we are not deploying technologies that way.
In manufacturing, David Johnson, Vice President, New Model Quality and production Engineering for Nissan North-America, observed just how complex today’s automobiles are, with approximately 30,000 components; and also believed that big data analytics and machine learning is able to assist with both present performance and in order to predict future performance. He mentioned that with virtual reality, Nissan is able to make decisions on new product announcements before anybody is able to see the physical new product, and specialists on the manufacturing unit could be brought in to help determine the process.