AI’s double-edged sword of progress
Expert: Getting automated out may open more, better jobs
There’s this famous book, titled ‘The Age of AI and Our Human Future,’ based on the beliefs and views of Henry Kissinger (former secretary of state and national security advisor of the United States), Eric Schmidt (former CEO of Google), and Daniel Huttenlocher (an American computer scientist and the inaugural dean of the MIT Schwarzman College of Computing) about artificial intelligence (AI). In the book, all three argue that artificial intelligence will soon make an impact on all human affairs. The sentence is simple and short, but strange and somewhat frightening. Automatically, scenes of famous works of the world cinema and literature on the subject of robots are evoked in the mind; scenes of humanoid robots that are made by humans to serve him, but then spin out of human control and become the enemy of men; A powerful enemy devoid of true compassion. Artificial intelligence has crept into human life for several years and has gradually shown its extraordinary abilities to today’s generation. Recreating the voices of singers who have long passed away or simulating the movements of our grandparents from their photos are just simple examples of the use of artificial intelligence that go viral every now and then. Today, there is almost no field that has not been taken over by the technology of artificial intelligence. Artificial intelligence has actually filled the shoes of humans and is gradually replacing them in all activities. Artificial intelligence has upgraded the robots from their initial form, which ultimately played the role of simple workers in factories, to designers and managers of production lines. This has raised more alarms in various industries, making people seriously concerned about keeping their jobs. With the penetration of artificial intelligence in different parts of human life, the concerns have gone even further; The concern is that robots and artificial intelligence will soon take over all the things that humans used to do and somehow replace humans. We are talking about very intelligent robots or machines with the ability to analyze and manage issues and make decisions like humans do or even much better. “Superhumans” is a term that is used to refer to such intelligent robots; Robots that have the ability to invent and create by themselves, a trait that has been unique to humans among the creatures of this planet due to their intelligence and thinking power. Now, what will happen if someday these smart machines gain free will? Is it possible that one day intelligent robots will take control of the world and rule over humans? At this stage, no one knows. For now, this theory has proven more substantial for movies, but the question is how far will robots and artificial intelligence go and to what extent will they enter the social and emotional relationships of humans. Maybe the production of those works of art about the future of robots is also a result of this human concern. In order to find an answer to these questions and concerns, Iran Daily conducted an interview with Hamid Zarrabi-Zadeh, the chair of the Computer Engineering Faculty of Sharif University of Technology. He has also been in charge of organizing the annual international Rayan AI Contest on artificial intelligence and programming.
By Sadeq Dehqan
Staff writer
IRAN DAILY: Due to the rapid spread of artificial intelligence in various sectors and its replacement of human activities, some people are worried that the continuation of this process will lead to widespread unemployment of people from their jobs. How serious is this concern?
ZARRABI-ZADEH: The risks created by the development of artificial intelligence are really serious in some fields. Today, in some cases, artificial intelligence has the ability to practically eliminate jobs, for example, in the fields of medicine and especially general practice, now artificial intelligence provides sometimes more detailed recommendations on the treatment of patients, so that people gradually no longer need to go to a general practitioner. Or, in the case of driving, self-driving cars using artificial intelligence and computers will soon take over a major part of intercity driving in countries like the United States, and human drivers may no longer be needed in those roles. So, the problem is very much real, and soon, we will see it leave its mark in other jobs as well.
To be sure, unemployment is a major concern. What solutions can we come up with to mitigate its effects, given the situation?
Well, the thing is, such global changes are inevitable as we’ve seen it before in other cases. We’re moving from one era to another. Remember when online taxi services like Uber in the US or Snap in Iran were launched? Everyone thought they would put round-the-clock call taxis out of business, but what actually happened was that many drivers from those agencies started working for Snap instead, and nothing bad occurred. Just the opposite, online taxi services ended up creating more job opportunities for the population at large. In fact, it was a step up for many people in the industry. What is more, transportation has become more accessible for everyone.
With technology advancing, jobs are shifting. The creation of jobs at a higher level is generally a boon for society because they bring with themselves new facilities and quality of life improvements. However, it can be tough on some jobs in the short term. Some jobs might not be able to persist in their previous forms. So, after an interval, either those jobs will adapt and survive at a higher level or their practitioners will find other positions to work at.
Artificial intelligence might have some negative effects in the short term for some jobs, but society will catch up quickly. AI is just a technology, and like any technology, it’s designed to make our lives easier. But let’s not forget that this doesn’t mean using a piece of technology more will necessarily make our lives better.
How does AI differ from other technologies we’ve seen before?
Well, other technologies have had an impact on our lives as well, but the case of AI has been different. In previous years, when robots were placing many workers in factories, those workers were first out of jobs but then moved on to higher-level roles, and machines took over the lower-level tasks. The concern with AI is that it’s gradually taking over higher-level jobs, where human intellect was once demanded. Of course, this would happen slowly, but if it does, humans would not be needed for many things that they took upon themselves to do, and this is where the threat is felt the stronger
Can you tell us more about the competition?
The first stage of the competition, or the preliminary stage, spanned for about a month from October 17. In this stage, participants competed with each other online. In the end, the top 150 teams moved on to the next stage. This stage was attended by around 10,000 people from 57 countries.
The second stage or the selection stage of the competition was also held online in November. In this stage, three questions related to the topic of trustworthiness in artificial intelligence were raised, and the top 15 teams with the highest scores were selected for the final stage. These teams moved on to the final stage to be judged by local and international judges.
What risks do AI’s imitations pose for social and emotional relationships?
In this area, too, AI has its pros and cons. Take ChatGPT, for example. It’s a chatbot that simulates human-like conversations. Many people are currently using it to meet some of their emotional and relationship needs. Studies have shown smart chatbots to be helpful here and there, including in cases where people need to vent to somebody or a therapist.
But there’s this concern that AI chatbots can pick up on the vast amount of information available on the internet and use it to generate conversations, which raises the risk that their conversations might veer off in a particular direction or be manipulated to serve specific purposes. Now, there are models that are specifically designed to fuel such misinterpretations and are called adversarial or hostile models. These models start disrupting the learning process of machine-learning models, just like a cyber-attack or hacking attempt. Such actions can cause other machines to spiral out of control and produce incorrect outputs because they don’t pick up on it and just act based on the data they’ve been fed. And so, if the learning process is disrupted or incorrect data is inputted, it can lead to bad advice and cause problems.
In the international AI program we ran last year, we actually set aside a major part of the competition to focus on the topic of trustworthiness in learning. The idea was to figure out how to ramp up the reliability of AI and take steps to ensure that these hostile attacks don’t have a significant impact on machine learning, or at least minimize their impact. This is what’s known as trustworthy machine learning, and it just so happens that the main theme of this year’s AI competition was also centered around trustworthiness in learning.
Humans have witnessed various eras of history, such as the Industrial Age, the Age of Communication, and so on. Will the impact of artificial intelligence in the coming years be so significant that we can call it the Age of Artificial Intelligence?
The arguments for it have been very seriously raised, and it is said that this era has already begun, and we are currently in it. Given the rapid progress being made in the field of artificial intelligence, we can give this name to the new era.
The progress of artificial intelligence in recent years has been beyond expectations, and now the talk of superhumans has been brought up, meaning we may be able to make machines that surpass human capabilities. Now, there are claims that machines have been built that can perform at the level of some human functions, such as robots or machines that can speak and produce new and varied sentences and expressions like humans. Until now, machines could only respond based on what they saw or heard, but now generative models produce things that did not exist before, even in the fields of art and music. In fact, machines have become creative, doing things that were previously exclusive to humans. Creativity has been transferred to machines, and machines can perform tasks similar to humans.
Currently, the subject of global studies — the results of which have not been leaked so much — is the topic of powers beyond human capabilities or superhumans. They can be more dangerous.
We may reach a point where humans may be under the control of machines, and intelligent robots may dominate humans. In terms of the capabilities of robots and artificial intelligence, we are moving rapidly in that direction.
Can intelligent robots completely replace humans?
It’s not a definitive yes or no answer. To answer this question, we need to see what the main characteristics of humans are, as there is a lot of disagreement in this area. What distinguishes humans from other living beings? All things considered, perhaps machines will not completely replace humans, but they will be able to do many of the things that humans used to do. My personal opinion is that they cannot completely replace humans.
.