The era of AI-powered robots is already here

reporter-profile
권순우 2022.12.01 12:18 PDT
The era of AI-powered robots is already here
Martin Ford, author of 'Rule of the Robot' (출처 : Jae Kwon Son, Hyeji Jang )

An interview with Martin Ford, author of 'Rule of the Robot'
"In the future, AI will become a universal resource like electricity.”
Creative and skilled jobs will survive the emergence of AI

What comes to mind when you think about ‘Artificial Intelligence’ (AI) or robots? Though everyone will have their own answer, many millennials and even younger boomers are likely to recall the classic Sci-Fi movie ‘Terminator’.

Martin Ford, the author of the 2015 New York Times bestseller, ‘Rise of the Robots’, notes that “It is a common media mistake to associate concerns about robotics and lethal autonomous weapons with films like ‘Terminator’.”

The idea of a war between humans and robots has long been a core theme in Sci-Fi. Currently, however, given the limitations of modern AI, it’s hard to imagine AI having the free will necessary to attack humans. Many experts have written off the notion as far-fetched.

In Ford’s opinion, the 2017 short film ‘Slaughterbots’, which depicts hundreds of miniature drones attacking the US Capitol Building, is closer to reality. In the film, the drones are fitted with facial recognition technology that allows them to target and attack specific politicians in an unmanned kamikaze-like mission.

The film was produced by Professor of Computer Science, Stuart Russell and his team at UC Berkeley, and featured solely AI technology that already exists and is in use.  Russell made the movie specifically to warn the public about the ways in which AI could be deployed as part of lethal weaponry.

"This isn't a future concern," Ford said of the film. "We should be more concerned about what humans will do with weapons (like drones) that aren't smarter than an iPhone but aren't afraid to identify and track targets," he emphasizes.

Ford's new book, 'Rule of the Robot,' is a realistic depiction of how AI can advance and what that means for the future of robots and humanity.

What follows are excerpts from The Miilk’s interview with him.

Martin Ford (출처 : Martin Ford )

"In the future, AI will become a universal resource like electricity"

The four books you've written so far are all related to AI. What interests you about the subject? Why this specific topic? (His 2015 book, Rise of the Robots, was named a New York Times best-seller.)

"I've always been interested in artificial intelligence because my undergraduate degree is in computer engineering and I worked in software development for many years." When I wrote my first book, 'The Lights in the Tunnel: Automation, Accelerating Technology, and the Economy of the Future,' in 2008, I began to think deeply about what advances in the field would mean for the future. Later that year, in 2015, I published 'Rise of the Robots: Technology and the Threat of a Jobless Future,' which expanded and continued on these themes. 'Rule of the Robots,' my most recent book, takes a much broader look at AI and what it means for society and the economy.”  

How has artificial intelligence evolved? Will it ever be smarter than humans?

"The field of artificial intelligence has seen dramatic progress in the last decade, owing largely to revolutions in Deep Learning" (or neural networks). This is a technology that dates back to the late 1940s, but advances in the techniques used, combined with the introduction of much faster computers and massive amounts of data, have resulted in many breakthroughs that would have been considered Science Fiction just a few years ago.

Language translation, image and speech recognition, the creation of fabricated images ("deep fakes"), and large language models like GPT-3, which are making remarkable progress in conversational AI, are some examples. Physical robots powered by AI are also becoming far more agile and capable. Despite recent advances, we are only scratching the surface of what is likely to be possible in the future. The goal is to create an AI system capable of thinking at the level of a human and beyond. Most AI researchers believe this will be possible and likely in the future."

It's intriguing to think of AI as a "new electricity."

"The main theme of my book, 'Rule of the Robots,' is that artificial intelligence (AI) is evolving into a general-purpose, systemic technology; eventually, it may be as ubiquitous as electricity." Electricity affects every aspect of our lives and we are completely reliant on it. It is a universal resource on which we rely in a variety of ways. I believe the same is true of artificial intelligence."

What infrastructures are required for us to start using AI as broadly as electricity? 

"Cloud computing is currently the most important infrastructural component for AI: the massive computing and data centers run by companies such as Google and Amazon." The majority of powerful AI applications run in the cloud, and cloud computing is now used by so many businesses and organizations that it is practically ubiquitous.

Of course, all of the devices we use to access the cloud, such as smartphones and computers, are an important part of this infrastructure. AI is increasingly being used locally on these devices, as well as in self-driving cars and other applications."

"Tesla's humanoid bot is only for promotional purposes. It’s hard to match the manual dexterity of humans"

Tesla recently introduced a humanoid robot. How did you feel about this?

"I feel that it is a bit of a gimmick. If Tesla wants to build a flexible robot for use in manufacturing, there is no reason to give it a humanoid form. Other configurations would have been more specific.

When the robot was first introduced, Elon Musk claimed it would be able to go to the store and do your shopping for you, but this is a science fiction fantasy. It will take a long time before we have humanoid robots with that kind of capability. Ultimately, Musk is vastly over-promising on what he can deliver.” 

Some would argue the humanoid bot did have a specific focus: using hands. 

Currently, robots cannot match humans when it comes to dexterity. However, the field is advancing rapidly and, within the next 5 to 10 years, we will likely have robots that can grasp objects as effectively as people can.

This will have a huge impact on employment in factories and warehouses. However, I am a bit doubtful that Tesla’s robot will be the best example of this."

In the past, you have criticized Tesla's AI-based autonomous driving technology. How far do you think this technology has come now?

“I think eventually, we will have true Robo-taxis that can operate on nearly any public road. However, we are not there yet, and the journey has taken longer than most people expected.  Lately, even Musk has backed off some of his overly optimistic projections. 

This technology may not be available for another ten years. In the meantime, we are seeing functional self-driving cars that can operate in very limited areas or on specific routes."

- Amazon's push for home automation is also heavily reliant on AI. When will home robots become a reality?

"I think it will be a long time before we have truly useful home robots." Building a low-cost robot that can move around the house and manipulate objects, such as bringing you a drink, is extremely difficult.

Even if we can do it technically, it will most likely be far too expensive for the consumer market at first. In essence, it will take a long time for technology to advance and costs to fall."

Tesla's Humanoid bot (출처 : Tesla Youtube)

Smart factories: the era of 'Robots as a Service' is fast approaching

Many manufacturers are relocating production back to the United States as a result of 'reshoring'  Naturally, this has increased interest in 'Smart Factories.' How advanced are modern warehouses and factories in terms of automation?

“Factories and advanced warehouses, like those used by Amazon, are already highly automated. All the truly rote, repetitive tasks have been automated. The jobs that are left generally require human-level dexterity, visual perception, or problem-solving. So far, robots cannot match humans in these areas.

This, however, is changing. Many well-funded startup companies, such as Vicarious and Covariant, are working on developing more flexible, dexterous robots.

Big companies like Amazon are also working on this. In 2018, Jeff Bezos predicted that robotic hands would likely match human capability in about ten years. When we reach that point, things could change quickly, with many factories and warehouses approaching full automation.

Another issue is that industrial robots have traditionally required extensive and time-consuming programming. In other words, robots are less adaptable than humans, making it much more difficult to re-orient them to tackle a new task. Again, deep learning AI-powered technology is changing this."

Various industrial robots were introduced at MODEX, a supply chain exhibition held in Atlanta earlier this year. (출처 : Soonwoo Kwon)

"Any job that requires creativity or skill will survive."

As AI develops, which field will it impact most? Where will humanity benefit most from AI?

"AI will have an impact on every industry and sector of the economy." It will increase business efficiency and productivity while lowering the cost of goods and services. In other words, it will contribute to the creation of an abundance of the things people require to thrive, which can be an important part of the solution to global poverty.

I believe that the most significant impact of AI will be to increase human intelligence and creativity in fields such as science, medicine, and engineering. DeepMind's recent development of 'AlphaFold,' which uses AI technology similar to what is already used in 'AlphaGo,' to solve the problem of predicting how protein molecules fold after they are fabricated in cells, is one example of this.

Protein molecules fold into complex 3-dimensional shapes almost instantly after being created, and the shape of the molecule determines its function, so predicting how they will fold is critical to biochemistry and medicine research. Scientists have been working on this problem for 50 years with little success. DeepMind was able to solve the issue largely in about two years, and now has a catalog of the shapes of all the protein molecules important in human biology. Many scientists believe that this will have a transformative impact on medical research and drug discovery.

All of this should lead to major breakthroughs that will propel progress in a variety of fields, including clean energy, transportation, medicine, and so on. AI will be critical in solving the major problems we will face in the future, such as climate change and global poverty."

The most commonly asked question is often what impact AI will have on jobs. In your opinion, which jobs will be created, and which will disappear? Will robots be able to replace humans in all roles?

"In the long run, once AI matches or exceeds human intelligence, almost nothing may be impossible to automate." AI is already creating art, which most people consider to be uniquely human.

In the short to medium term, I believe there will be three primary areas where human workers will be relatively safe. First and foremost, jobs that are truly creative. For the majority of these workers, AI will be a supplement to their skills rather than a replacement.

Secondly, jobs that require deep interaction and relationship-building with people. Algorithms are already beginning to interact and converse with people, but I think it will be a long time before machines can build truly meaningful relationships with people.

Lastly, skilled trade jobs requiring a high level of mobility, dexterity, and problem-solving in volatile environments, such as electricians and plumbers, will be difficult to replace. Automating jobs of this nature would necessitate a Science Fiction-level robot on the level of C-3PO from Star Wars. It will be a long time before we have such technology."

Can AI solve the ongoing problem of social inequality? You mentioned implementing a universal basic income, as proposed by Andrew Yang. Do you think this is feasible?

"I believe AI will worsen rather than improve income inequality, so we must consider policy solutions." As AI begins to encroach on nearly all jobs, we may need to develop entirely new social contracts in the long run.

As a result, I believe that a universal basic income is an important concept to consider for the future. If a large portion of the population is disenfranchised from the labor market, we will have to find another way to provide people with an income so that they can survive economically and fulfill their role as consumers. This is critical to the market economy's success."

"The emergence of Deep Fakes is a threat to democracy”

What is the greatest negative impact that AI currently has or could have on mankind? 

"In my book, I discuss a number of significant risks. AI is already being used for surveillance and invasion of privacy, particularly in authoritarian states like China. It can also be used to generate "Deep Fakes," which are extremely high-quality fabricated pieces of media such as audio recordings, photos, and even videos. Deep Fakes could eventually be used to influence elections or otherwise undermine democracy.

AI can be used to build autonomous weapons capable of killing people without any authorization from a human who is “in the loop.” It will result in the automation of millions of jobs, and we will have to find a way to adapt our economy and society to that reality. If AI becomes truly powerful, we may face a threat from machines that are smarter than us and may act in ways that are harmful to us.

Nonetheless, I believe the advantages outweigh the risks. Our goal should be to maximize the advantages we can gain from AI while minimizing the risks. In many cases, this will necessitate some level of regulation of AI applications."

(출처 : Shutterstock)

In your book, you compare the future of AI to the films 'Star Trek' and 'The Matrix.' How do you envision the future of AI? (As the most optimistic future model, Ford cited Star Trek's vision of a world that values intrinsic humanity over economic gain. 'The Matrix,' on the other hand, depicts a more dystopian world in which people's rights and opportunities have been taken away by AI, forcing them to seek refuge in an alternate reality.)

"I hope the future is like Star Trek, but I'm afraid the default is more like the Matrix." As I stated in the conclusion of 'Rule of the Robots,' I am concerned that AI will make the world much more unequal and difficult for average people to thrive. This is already happening, as evidenced by the popularity of the television show "Squid Game." If things worsen, many people may decide to spend a large portion of their time escaping into virtual reality or using drugs."

What do you hope to achieve with your new book?

"AI will have a profound impact on all of us." It will become one of our most important technologies and a major force in shaping our future. It will eventually have an impact on both society as a whole and every individual who lives within it.

Many people's jobs will be eliminated or drastically altered as a result of AI, and there are significant risks and opportunities associated with the technology. As a result, I believe that everyone should have a basic understanding of artificial intelligence and its implications. That was my intention when I wrote 'Rule of the Robots.' It is not a technical book, and I hope that anyone who is interested can read it."

영화 '메트릭스4' (출처 : 워너브라더스)

회원가입 후 뷰스레터를
주 3회 무료로 받아보세요!

단순 뉴스 서비스가 아닌 세상과 산업의 종합적인 관점(Viewpoints)을 전달드립니다. 뷰스레터는 주 3회(월, 수, 금) 보내드립니다.