A team of researchers from Boston University published a survey, and I received an email about the same-titled “Should artificial intelligence replace journalists, judges, hiring managers?” One may wonder why humans want artificial intelligence to have such key roles. But as it turns out, this might eventually happen.
The survey asks if people want to replace journalists, trial judges, hiring managers, spiritual guides, and religious leaders with AI. People are still warming up to the idea of AI in our lives, so there isn’t much acceptance for now. However, even today, nearly 30% of people say it is a good idea to replace journalists with Artificial Intelligence.
James Katz, Boston University College of Communication’s Feld Professor of Emerging Media, pointed out this trend. Prof. Katz is also the co-editor of the book Journalism and Truth in an Age of Social Media. We reached out to him for an interview and got some interesting takes on AI replacing human jobs in the future.
But before we dive into it, here are 6 things you should know from the Boston University survey on AI replacing humans.
6 Highlights: Boston University survey on AI Replacing Human Jobs
- Journalists losing to AI: This is one of the most profound findings of the study. Younger people aged between 18-34 years old want AI to replace journalists. This is because people want their news to be free from bias.
- Spiritual leaders to remain human, for now: Over 70% of the people in the survey either think it is a bad idea to have an AI spiritual guide or straight out dismiss the possibility. The results remain pretty much the same for religious leaders.
- Younger people aged 18 to 34 years are more open to the idea of AI replacing all jobs.
- Men are almost 10% more open to the idea of AI replacing jobs than women.
- The pattern emerges: The survey suggests that even today, at the beginner stage of AI advances, close to 10% of surveyed people are open to future possibilities.
- What would we do if AI replaced all jobs? Well, that’s something Professor Katz explains in the interview below. Do read to the end, and leave your thoughts in the comments.
In conversation with Professor James E. Katz
Would you like to present an outline of the survey and its findings? Did these findings surprise you?
We began this survey because we wanted to understand the general popular feeling towards the idea that artificial intelligence would replace human beings in a variety of different roles. And, of course, most people understandably don’t know much about artificial intelligence or how it works.
But we wanted to probe the public’s general sentiments about how they would feel about this phenomenon taking place. And one of the dimensions that we wanted to understand is whether the role that the person takes makes a difference in how they feel about having them replaced by an AI agent.
What we found is that, surprisingly, about a third of the general U.S. population is favorable to the idea that different roles would be replaced by machines. Some of the other roles we asked about, for example, a trial judge, that I think you could see possibly being replaced by a machine that would not be influenced by passion.
But by and large, people are about 30% favorable to seeing people replaced. 30% were neutral, and 30-40% were opposed. And within that, we found journalists to be the most vulnerable to being replaced by AI.
The survey says that men are 10% more receptive to artificial intelligence replacing jobs than women. Why is this pattern there?
Well, first, we found a lot of variation by age, and I don’t think that it would surprise anyone to hear that young people are much more favorably inclined towards replacing people with AI. After all, they interact with machines, i.e. mobile phones and computers, all day long.
Whereas older people are much more set in their ways and, generally speaking, have more fear of change, especially technological change. But within that, the gender dimension is also interesting and valuable to understand.
Although we didn’t delve more deeply into why men are more favorable to AI replacement of people than women, I think gender is a fairly consistent predictor variable of people’s openness to artificial intelligence specifically and technology generally.
A common stereotype is that men are better with machines than women. And although there are huge variations, and in many cases, women are much better with computers than men. On average, statistics show women are less inclined to technological innovations on the computer front than men.
So this is brought out, consistent with other studies about gender differences in the appeal of technology. It is brought out in the survey, and to summarize, men are significantly more favorable to having professional roles substituted by AI agents than women.
There was an observation where there were two different roles. One was that of a spiritual advisor, and the other one was that of a leader of a religious congregation. How are these two different, and how did you help people differentiate between them?
I’ve written a book called Magic In The Air about mobile phones. In that book, now a couple of decades old, I explored the spiritual dimension of mobile phones. I wanted to see how some people find a deep emotional investment in their mobile phones, and treated them like a friend, like an advisor.
Like a mystical source of being able to connect with worlds far beyond the ordinary nature of everyday activities. Some of my students have followed this line of research. They point out that religion has a specific set of guidelines or norms or beliefs. It tends to have an organizational structure, and of course, religion is a spiritual dimension of people.
On the other hand, a lot of people have, what might be considered, a different dimension of spirituality or belief, and that may not be organized like religion in the traditional sense. Although it can and does overlap. So we wanted to see whether there’s a difference in people’s perception if they conceive a thing as spiritual vs religion.
Here in the U.S., participation in organized religion has been declining over the past several decades. Whereas, according to surveys, spiritual activity and spiritual beliefs remain high. So here we see a distinction between the two forms of engagement. With forms of life that extends beyond everyday reality, and so our purpose was to do some initial exploration between different domains to see to what extent they’re identical and to what extent they overlap.
And this then connects back to that research interest of mine. Trying to understand the relationship between ordinary reality as we experience it and our reliance on different forms of communication technology. As I said originally, mobile phones, and currently artificial intelligence.
There are interests within organized religion and spiritual communities in using AI and robots as forms of delivery or inherence of messages, capsulation of voice to connect to the higher levels of human experience. In Europe, they have a robot priest with similar AI kinds of supplements to ordinary religious and spiritual practices.
Robots and religion have a historical connection
And, we go back to ancient Greece; there, some of the very earliest machines invented were included in temples. Machines that, in exchange for a coin, gave some sort of spiritual incantation. A mechanical bird would fly, or mechanical eggs would be produced.
So I’m pleased to continue this tradition, now in a social scientific way, to explore how people interact with these devices to connect with higher levels.
People collect and curate the data that we use to train AI. So how does people’s bias not creep into that database and haunt what AI is creating?
I think it is important for those designing AI algorithms to have a diverse community upon which they develop their algorithms. Now in today’s environment, we’re very sensitive to certain demographic qualities. For example, here in the U.S., research is done, say, on people from a wide ethnic male community, would not be extensible to non-white non-male communities.
Another example is that AI may be trained on extroverts, because they’re the ones who sign up for this kind of testing opportunity, whereas introverts probably don’t. Therefore it is biased towards introverts.
It is important to train these systems with a meaningful set of criteria. So in the long circuitous answer to your question, it is important that we have accurate training data to establish these systems. And that remains a huge challenge for the field.
There’s another side, which is that it is important that we understand once these data are gathered and given to the machine. The machine would often use self-learning algorithms and take off with the opportunity to train itself.
So, in other words, we’ll have a black box that’ll make decisions and make choices. It may be trained initially on a fair or unfair information, but then it takes off and begins to operate in a way that is not viewable, not perceptible to the human designers or evaluators.
And that is the second-order level of concern that we need to address. As one that, compared to the bias question that we began with, is far less understood in ramifications.
According to a study, journalists working in tandem with AI are earning the reader’s trust more. Will AI replace journalists or complement their work?
I think AI has a fabulous opportunity to counteract certain forms of bias that creep into journalism. And I think one of the low-level interventions is already available on some word processing platforms. For example, using AI to identify gender bias terminology.
There is another form of bias that is harder to capture with AI or any other system. These are conceptual biases that journalists encounter. For example, there may be a bias towards the underdog for example. In the past, maybe too much attention was paid to the elite in writing, and today “too much” attention is being given to other groups. So is this a bias or not a bias?
Is this just how the world proceeds? We call these zeitgeists, spirits of time that lead people to have blinders they’re not even aware of, and in the future, they’re called out for the biases that they were unaware of at the time. So these are higher forms of bias that we need to make our way through the world.
We can’t absolutely define all of our terms every time we use a term. For example, we can’t make sure we’re identifying and then therefore purposely excluding people of one category while talking of another category. So these are high-level biases, and AI will be more severely challenged in trying to counter these. It is so because AI can’t create awareness of bias at higher levels out of thin air.
They have to be trained & developed, and that implies, once again, that trainers and individuals that create systems have to be aware in the first place. That’s almost god-like powers that are rare, if not unavailable to people.
As an expert on the subject, what would be your opinion on AI actually attempting to replace human jobs?
Well, traditionally, as people get older, they become less technologically enthusiastic and more stuck in their ways. These ways can be good, bad, or indifferent. So we can expect those young people who are so enthusiastic about technological change replacing jobs to be more like the older generations that they currently like to think of themselves as so different from.
Many years ago, I was part of a panel that had to address the question of what people would do with all their free time in the year 2020. Because machines would be replacing all the jobs. There would be smart factories producing all the goods, and people would have nothing but free time on their hands.
And how could all that free time be absolved by entertainment and some other way because machines would replace all the dull and hard work, and leave people only with fun things to do. Well, I don’t know about you, but I’ve been as busy as I’ve ever been, and I hear complaints from people about how busy they are.
So I think the fear that we don’t have anything worthwhile to do as machines take over more and more of our lives is something that will never materialize. We’ll always be busy with important things to do, just as we are today.
People used to be busy with carbon paper typing and triplicate circulating of things and physically delivering envelopes. Today, we’re busy answering the huge crush of emails that come cascading down on us, and our problem is how can we deal with so many emails every day. Since you say, we’ll find new things to keep ourselves very busy in the future.
Professor James E. Katz observes artificial intelligence as it is today and for what it can be in the future. He has seen such developments in the past that he is optimistic about the future. According to him, future humans will find a way to keep themselves occupied. However, there’s also hope that people will contribute to something meaningful with that time.
Metaverse is another way to look at these developments. With Facebook now Meta’s aggressive push, we can expect VR headsets to become mainstream soon. Companies like Apple are also jumping into it, which makes things interesting.
These may be reasons to believe that artificial intelligence will play a major role in our lives. We’re more likely to become accepting of artificial intelligence as we become more welcoming to the idea of virtual reality. While both are two different concepts, they seem to be making strides of progress together.