In BriefJust as the agricultural, industrial, and cyber revolution led to fundamental shifts in how we live and work, so is the age of autonomy reshaping the very fabric of our society. Today, MIT's Daniela Rus offered her insights into what this new world will look like.
If you work in an ad agency, a robot is probably going to take your job. If you drive a taxi, or work for a ride-hailing service like Uber, a robot is definitely going to take your job — and will probably do so in the next couple of years. If robot autonomy doesn’t take your job (but just an FYI, it probably will) you are going to get paid less because of a robot.
We hear these facts on a near constant basis. New reports continually add to the list of the soon-jobless, saying that automation is going to steal work from lawyers, from writers, and even from the information technology experts developing and installing computer systems today. Estimates say up to a third of the workforce could become automated just over the next decade.
So, just how much do you have to worry?
To find out, Futurism’s Alex Klokus spoke at the World Government Summit with Daniela Rus, the Director of the Computer Science and Artificial Intelligence Laboratory (CSAIL) at MIT, about automation and the world that AIs are already creating.
Rus attended the Dubai summit to speak about the next great revolution: the autonomous revolution. Just as the agricultural, industrial, and cyber revolutions led to fundamental shifts in how we live and work, so is the age of autonomy reshaping the very fabric of our society.
Soon, Rus says, AI will be woven into every aspect of our existence. Yet she does not see this future as one of despondency and despair. Far from the economic upheaval and mass job loss portended by some, Rus envisions a future in which humans benefit from the fruits of AI labor.
When the topic of AI comes up, Rus noted that people frequently jump to the downsides: “Some people get very anxious and start asking about Skynet and when robots will start taking over their jobs. Well, I believe that everyone stands to benefit from AI,” Rus proclaimed.
When asked how she thinks governments can ensure that AI advances do, in fact, benefit humans, Rus stated that such guarantees will require grandiose shifts in how we think about education. Currently, Rus explained, we think that education has an end point, a point at which it doesn’t make sense to go any further. This needs to change.
“In the future, we will have a very parallel approach to working and learning,” Rus said. “We study, and study, and study, and then we say, ‘okay you learned enough, you can start working now.’ In the future, we will have to blend studying and working.”
This parallel approach is something that, Rus believes, governments and companies will need to figure out soon. Will companies be responsible for constantly training and retraining their workers so that they can keep pace with new facets of their work? Or will governments be responsible for this lifelong learning?
At the present juncture, there are no clear answers. This work hasn’t yet been carried out— or really even begun.
A Better Way of Seeing
As the world works to prepare itself for the age of automation, Rus says it must also prepare its citizens to have a different outlook on developments in AI and autonomy. Instead of jumping to killer robots and job loss, we need to provide alternative ways of thinking that allow people to understand the benefits of AI.
One such benefit can already be seen in the autonomous driving industry. Though level 5 autonomy is not quite here yet, and autonomous cars are not yet ready for public roads, Rus pointed out that driverless cars are already capable of driving people around places that have less congestion and hazards, such as parks and retirement homes. To whit, in her presentation Rus showed a short video of a golf cart-like car picking up an elderly passenger and driving her from her retirement home to meet her friend for lunch on the boardwalk nearby. This, Rus notes, is an example of a benefit of AI, and of how autonomous features can restore freedom and mobility to those who are confined to their apartments and houses.
“Notice that this car does not have a driver, but we should not see this is displacing work because this is a service that does not exist,” Rus said optimistically. (This incited a quiet laugh from philosopher Nick Bostrom, beside me, at her vision of a world where autonomous cars don’t replace human drivers.)
Rus also noted in her talk that health and medicine AI systems make a number of errors when analyzing medical images for signs of cancer — in 7.5 percent of cases, to be exact. Humans? They hovered around 3.5 percent. Yet together, humans working with AI machine systems lowered the error rate to just 0.5 percent. “Doctors will be able to offer the most advanced treatments by working in tandem with machines,” Rus said. She continued by noting that this is just one example which proves that AI will not replace human workers, but augment the work that they do. Some in the audience looked skeptical.
Rus reinforced her argument by noting that, when talking about job loss, lawyers are often right at the top of the list. However, she asserts these stories are hype and highly exaggerated. “Natural language systems can read entire libraries of books and provide information at just the right time, right when it is needed,” she said. “This doesn’t mean that machines are becoming lawyers. Machines are just changing the work that lawyers do.”
There is some truth to these claims. Take, for example, the legal technology company LawGeex, which created an AI algorithm that automatically reviews contracts. Automating such processes that are little more than paper-pushing has saved law firms a lot of money; however, the true advantage of these autonomous systems is saving attorneys’ time. Indeed, as one participant noted in an earlier round table talk on AI, “No one went to law school to cut and paste parts of a regulatory document.”
Overall, Rus says, she has two concerns: “One is about the quality of the job, and the other is about wages,” she said. “I am not concerned about whether we will have enough jobs, I am concerned about whether we will have enough good jobs.” For example, Rus described how GPS made it possible to become a taxi driver without knowing the maps and road systems. This lowered the entry point for work, but it also lowered wages.
In truth, the concerns don’t end there. Initially, robots will augment our work; yet if autonomous systems continue to progress, becoming faster and smarter, what is the guarantee that humans will truly have a place? Of course, there is none. Yet our social, political, and economic systems can evolve. And so can we.
Machines have the power to support us or to hurt humanity; it is up to us to decide how we interact with them, how we govern them, and how we set the rules for machines — as well as the necessary legislation to govern humanity — to ensure that all of these advances are for the greater good.