Technologies of Tomorrow
Skyline at Night - Shanghai, China
I’ve always had a fascination with science fiction cinema. The stories and characters would catch the interest of my younger self, but what kept me hooked was behind these elements. From the older Star Wars movies, Pixars’ animated Wall-E, all the way to the new Blade Runner 2049 and all the TV series in between, I would pay close attention to the technology being displayed and utilized throughout both the background and the characters. Some of these things would blow my mind, causing me to pause on a frame and examine a uniquely styled piece of architecture, vehicle, or gadget, wondering how it could possibly function in today’s realism. It was this fascination which kept me hooked on the sci-fi genre, and even inspired some of my academic studies and past experiences.
​
As someone who has been deeply involved in technology since a very young age, I’ve kept close attention to digital elements involved in our lives, as well as those to come in the near or distant future. I closely relate the progression of technology in today’s world back to what I see in science fiction, examining the benefits, consequences, and externalities they possess, and how the combination of so many shape our society. What motivates my curiosity and interest in all this is the fact that some of these fictional depictions of technology are much closer to reality than we think.
​
At this point in time, I’ve come to gather a considerable amount of knowledge on two separate, yet connected technologies: artificial intelligence and drones. My credibility on these topics stems from not only independent and scholarly studies, but also has been heavily influenced by past experiences I’ve had and reflected on. The most significant of these experiences took place throughout my first year of college, during which I lived and studied in the international mega-city of Shanghai, China, where I attended New York University’s international campus. This was a powerful place to be with my interests and questions in mind, as China takes a multi-mile long lead when it comes to societies that function by coexisting with and by utilizing the prospect of innovative technologies. Witnessing and studying this dynamic first hand in China was an invaluable experience. Class projects required me to develop early-stage applications for drone technology, and even pitch their functionalities and business plans to investors and critics in the famous Shanghai World Financial Center. Through a colleague of mine, I had the wonderful opportunity to travel to Shenzhen, a city recognized as the silicon-valley of the eastern hemisphere located just north of Hong Kong. This is where I visited DJI’s headquarters, the world's largest commercial drone manufacturer, and toured their drone development and robotics division. I’ve also attended various seminars, each involving some of the most technologically influential people in Asia. These endeavors have allowed me to experience my fascination with innovative technologies through theory, development, and action.
​
The remainder of this article will explore these two technologies, artificial intelligence and drones, sharing what I’ve come to understand, as well as how I expect them to be used. By discussing their capabilities and applications, I hope to share my excitement about the promising future of these technologies and the incredible impact they will have on our lives. However, because of the potential for these topics to be disruptive and possibly catastrophic, I do have some questions and concerns. By addressing both these expectations and concerns from an interdisciplinary perspective, I hope to shed light on what may very well be the reality, and problems, of tomorrow.
New York University - Shanghai Campus
Artificial Intelligence
The Prerequisite for Technological Evolution
It is undeniable that modern technology has greatly assisted mankind in the process of advancing society to greater levels of productivity. Furthermore, such technologies have allowed for people to evolve as a species. From the unimaginable era of the stone age and throughout the history which followed, technology has served as a tool to allow us to reach unseen heights of creation, cultivation, and comprehension. In the present, this tool comes in limitless forms; from satellites to smartphones and from oil rigs to automobiles. The results of these technologies structure our world today. Many even argue that the technology we possess is a strong indication of our species’ intelligence. However, we have reached an inevitable plateau, where ‘next-generation’ technologies seem to be stuck on the horizon; where the promising universal future of flying cars and holographic phones seems only to appear in science fiction. Our current dilemma is that the technologies we utilize must reach higher levels of capabilities, specifically those which can mimic humans in the way we make decisions. These decisions are any machine's output, and any output requires an input. Modern technology output is limited because it does not possess the ability to think humanly or rationally. This may not matter in the case of using a hand-held power drill. However, in the case of anything autonomous, such as a self-driving car, it is the most important factor. It is widely understood by professionals that in order to achieve the next level of technological evolution, our machines themselves must become more intelligent.
​
This is where the term ‘artificial intelligence’ is coined. The ability for machines to possess such effective levels of decision making would simultaneously propel any process into new standards of effectiveness. In my opinion, Artificial Intelligence (AI) is best described as an application. While AI is certainly a technology independent in its respective field, it is a technology that is applied to other technologies, systems, or processes. This is why when people hear the term “AI”, it is often being phrased in conjunction with another label or title, such as “AI in our smartphone applications”, or “AI in Medicine”.
​
In simplest terms, artificial intelligence is a highly advanced computation of information. The result of this computation is the ability for machines to think and work like humans, resulting in machine intelligence, in contrast to mankind's natural intelligence. The common goal for this technology is for it to possess the following four abilities: thinking humanly, thinking rationally, acting humanly, and acting rationally. AI does this by utilizing an enormous amount of inputs, and by having powerful enough computation power to produce an output based on those inputs. This is where the subject ‘big data’ is utilized, and as long as the computer is sophisticated enough to process the information, the more data points, the more productive and precise the output will be. Another term that one may come across when researching AI technology is ‘machine learning’. The label seems self-explanatory, but what it fully refers to is the method in which we can create these AI systems to learn from the results of any given task. The results are then included into the input of the system, so the next time the task is approached, previous outcomes of the same (or similar) task are accounted for. Computer engineers have been taking this a step forward recently, running simulated environments in computer software where the AI system can perform a given task tens of thousands of times over, learning from its mistakes and accomplishments, storing all gained knowledge into what is called a ‘neural network’. This method of machine learning has been experimented with various scales of intensity and complexity. The short clip below is an example of a less-intricate machine learning simulation experimented with the notoriously addictive and viral smartphone game, Flappy Bird.
IBM's Quantum Computer
I/O Structure in a Neural Network
Before explaining what is going on here, it is important to know that the goal of this Flappy Bird game is to fly between the green pipes for as long as you can. If your bird hits a pipe, the game is over. Hobbyists alike have simulated the environment in a machine-learning system, allowing the computer to attempt multiple trials simultaneously. Each trial is known as a ‘generation’. The AI learns from its mistakes and records the behavior of trials that progress further than others into a digitized neural network. The end goal of this process is to have an AI that embeds enough rules and behavior into this network so that it can autonomously soar through the level uninterrupted and further than any natural player ever could.
This self-teaching method used in artificial intelligence is so much more meaningful than simply mastering a smartphone game. This is merely an example for the process behind the scenes. In addition to being an insightful comparison, the following video clip demonstrates a higher level of complexity in a machine learning system.
da Vinci Robotic Surgical System
This project authored by Samuel Arzt posted on YouTube demonstrates a small 2D simulation in which cars learn to navigate through a course without any human input. Similar to the Flappy Bird experiment, Samuel Artz programmed the cars to use a neural network to learn from previous generations. By the fifteenth generation, one of the many cars finally maneuvers through the course without coming into contact with the walls. At the end of the video, and 31 generations later, we can see half of the cars successfully made it to the end.
​
I include this video for two reasons. First, this experimentation with neural networks in simplistic objective-oriented games further demonstrates how machine learning is depended on in the process of building intelligent systems. Second, this experiment by Samuel Artz can be closely related to the prospect of autonomy. The 2D cars in this video are built on infrastructure that devices in the future will depend on as well. Advancements in artificial intelligence are the prerequisite for advancements in other innovative technologies. It is also the backbone for autonomy - arguably the most critical ability that advanced technological devices must utilize.
​
Understanding the concept of the logic behind AI and qualifying it as an applicative technology, you could begin to imagine what fields it could evolve. In the context of societal evolution, let’s touch back on the topic of AI in medicine.
​
Reducing Human Error
​
Applied to the medical and healthcare industry, artificial intelligence is reinventing and reinvigorating modern healthcare through the use of machines that can predict, perceive, understand and act. One way in which it is being used for is to reduce human error. Every medical process, or any process rather, can become obstructed and deviated by human error. This phenomena can occur in any humanly task, and most of the time, is nothing too damaging or is accounted for properly. However when human error occurs during a medical operation, or during the diagnosis of a patient for example, the results can be devastating. In 2015 alone, misdiagnosis illness and medical error accounted for 10% of all deaths in the United States. (Daley, 2019)
Artificial intelligence applied to medical practices where human error often occurs, such as diagnosing a patient or during surgery, would not completely take over the process. While used synchronously with human professionals, AI is proven to reduce human error significantly. A great demonstration of this can be witnessed from a study published in the journal Nature. This study compared the results of mammograms analyzed by medical professionals with the same images but instead analyzed by an AI system that had been trained to identify tumors. Author Rachel Ellis states that “when the computer was asked to read images from nearly 29,000 women, the number of missed cancers known as false negatives fell by 2.7 per cent compared with when a single doctor reviewed the scans, while the number of mammograms incorrectly diagnosed as abnormal (known as false positives) decreased by 1.2 per cent.” (Ellis, R., 2020) The study concluded that the AI system performed as effectively as two experts working together to analyze the same images.
​
Rapid Diagnosis
​
Consulting a doctor in efforts to determine the source of symptoms or an appropriate therapy can sometimes be misleading. By utilizing big-data and machine learning technology, artificial intelligence could allow a digitized system of cataloging and matching data points together at speeds beyond what any doctor can achieve. Discussed in an article by Sam Daley, medical centers have already begun using artificial intelligence to diagnose potentially deadly blood diseases and cancer at very early stages. (Daley, 2019) According to the author, “doctors are using AI-enhanced microscopes to scan for harmful bacterias (like E. coli and staphylococcus) in blood samples at a faster rate than is possible using manual scanning. The scientists used 25,000 images of blood samples to teach the machines how to search for bacteria. The machines then learned how to identify and predict harmful bacteria in blood with 95% accuracy.” (Daley, 2019)
​
Precision Medicine
​
Arguably the most significant application of AI technology in the medical industry is the effort to pursue methods of better-targeted treatments. Artificial intelligence will allow healthcare to move from one-size-fits-all type treatments to more tailored therapies. The use of AI in this field is quite extraordinary, and according to Fabian V. Filipp, artificial intelligence technology “has an ability to predict novel drug targets, quantify ADME and toxicology, match drugs with targets and biological activities, model physicochemical properties, accelerate data mining, predict biological targets for compounds on a large scale, design new chemicals and syntheses, and analyze large virtual chemical spaces”. (Filipp, F. V., 2019, p 4) This new paradigm enables medical chemists to process and analyze billions of molecules in a virtual environment. This isn’t all though, as AI can also provide insight into drug mechanisms, establish biomarkers, repurpose existing drugs, optimize drug candidates, design clinical trials, and even recruit for clinical trials. (Filipp, F. V., 2019, p 4) Dr. Emilio Gómez-González agrees with this, adding that the “decoding the genome of each patient represents a very significant change from the existing model of averaged analysis of populations to an extremely individualized approach, for treating disease –in a new paradigm defined as “precision medicine”” (Gómez-González, 2020, p 8)
​
Just these few applications of artificial intelligence systems in the medical industry alone are quite literally life changing. While the technology of AI is still in development, its potential is promising, and its ability to maximize the productivity of respective tasks can be revolutionary. The ability for maximized efficiency in modern operations is increasing simultaneously with advancements in machine learning, and utilization of big data. By incorporating these computer-aided methods into medical practices, “we are moving toward the rapid expansion of the use of these tools in daily life of physicians, making each patient unique, as well as leading radiology toward the concept of multidisciplinary approach and precision medicine.” (Koenigkam, 2019) The future of AI in medicine may very well evolve society to unseen levels of healthcare and medicinal understanding.
Deep Learning with Doctus AI
Drones
The next level of transportation
DJI Mavic Pro
In simplest terms, drones are aerial vehicles that don’t have a pilot on board, also known as unmanned aerial vehicles (UAV for short). They can be autonomous, operating without any user input, or remotely controlled from afar. They’re mostly made of extremely light materials and utilize a number of propellers powered by on-board batteries. Today, civilian drones are becoming more mainstream than ever, and it is arguably a must-have device for photographers and videographers alike. At the moment, it is this reason that they are mostly used for. When equipped with a high-resolution camera and operated by a trained professional with virtual reality FPV goggles, the results can be truly extraordinary. Take a look at the three clips below.
​
​
.png)
​
Before drones, producers looking to get aerial shots for their project would have to hire registered aviation companies, using helicopters or planes with a camera on board. Since this would often come at a heavy price, the availability of drones has entirely put these companies out of business. Disruption of this level easily compares to the disruption seen by the inventions of ride sharing applications like Lyft and Uber, which stripped taxi conglomerates of customers, and sent the entire industry into panic.
It is this level of disruption which fascinates me most. Over the next few years, drones will completely displace and reshape conventional methods of business and consumerism, all due to the simple ability of being able to harness cargo with these devices. Whatever this cargo will be, whether its food, packages, or even people, the respective logistics infrastructure will also be left with a decision: to adapt, or fall behind.
​
Retail
Let's look at the E-commerce industry first. Companies like Amazon have begun experimenting with drones to deliver goods from local retailers and fulfillment centers. In 2016, the company made its first commercial drone delivery with their Prime Air system. Amazon announced its intention to roll out 30 minute or less deliveries to select Prime customers who use the service.
DJI FPV Goggles
Uber Eats Test Delivery
​
You don't need to look in the sky to realize that this hasn't become mainstream yet. At the moment, drones aren't even allowed to be flown through the airspace of basically every major city. Nonetheless, the technology certainly exists, though more development and testing is required before this becomes something we can completely depend on.
​
But once these devices become dependable, and regulations make room for drones to operate through populous places, will it completely revolutionize the E-commerce industry? An article by Adrian Lloyd discusses this, and points out how it would require a major overhaul of existing infrastructure. Supply chains for these companies would need to be modified in order to adapt to such a new standard. Lloyd explains that because these drones would replace carries in most cases, new systems will be required to match these products with drones. This could be of concern to E-commerce and retail companies of all magnitudes, they would need to anticipate this adaptation in order to minimize the expense of overhaul.
​
Lloyd brings up another good question, wondering if only the "big boys" in the industry will come out on top of this. Because the drones themselves as well as the modifications necessary to utilize these drones comes at such a cost, "will only companies like Amazon and eBay buy drones, while the smaller and mid-sized retailers get priced out of the market due to high costs?" Once the ability is there and reliable, rapid drone deliveries would undeniably be the preferred way to purchase. If customers can't get the same amount of satisfaction out of retailers who can't compete with this, will those retailers eventually fall out of the picture?
​
Food
Similar to retailers and E-commerce stores, applications of drones for the restaurant industry would cause a chain reaction of disruptive events. By creating a scenario in which this happens, the benefits brought by this technology, as well as the effects it would have becomes more clear.
​
Suppose some restaurant owner decides to purchase one of these newly released drones in an effort to convert their current food delivery infrastructure. It would be fairly simple. The drone would be autonomous, having the ability to navigate above a city's streets while safely avoiding obstacles. It would harness an attached cargo box where an order can be stored, with a short-term ventilation system to keep it fresh. Set by its newly proud owner would be the origin point, where the drone takes off from and returns to. This would be located somewhere accessible in the restaurant. Powered by onboard batteries, let's say the drone has a flight time of 45 minutes, and re-charges itself at the origin point when not in flight.
​
With a logically realistic scenario set in place we can assume the outcomes. The first and most obvious result would be a realized ability to deliver orders to customers at record times. The food is delivered fresher than it would be if received by a car. Most importantly, orders can now be placed from anywhere in the city and are no longer restricted to a small radius.
With the acquisition of exponentially more customers and the satisfaction of minute-wait deliveries, the restaurant begins to receive more and more orders. An economic decision is made to lay-off the old delivery employees and the restaurant continues to profit from its surplus of orders and mitigation of expenses. In fact, delivery orders begin to completely outweigh walk-in customers. It no longer makes sense to pay rent on a multi-thousand square foot lease, so the owner decides to downsize into a much smaller space that only serves via delivery, cutting down on staff and more expenses. The restaurant has reduced to its core, with minimal staff on payroll and much more business.
​
An article published on CREntrepreneur makes a powerful point that ties the use of drones in E-commerce and food delivery together. It will become a “game changer”, the author writes. “Why send a guy in a truck when a drone can do it faster, cheaper, and with fewer potential consequences?”
​
Surveillance and Security
​
Looking at the surveillance and security industry, the implementation of autonomous drone technology would bring significant innovation. Companies have already started working on this, and their projects in development demonstrate promising potential. Azur Drones, a French company founded in 2012, has developed a UAV they call Skeyetech. It is a fully autonomous drone solution for safety and security systems that work in even the most sensitive and complex environments. It can navigate through hazardous zones like fires or hostile intruders, all without jeopardizing an operators safety.
It is made to be used on large industrial properties that may hold sensitive information, such as data centers, nuclear plants, oil rigs, and cargo ports. Once configured, the drones fly along previously programmed routes, using high-definition cameras with thermal optics and infra-red to detect trespassers or anomalies. If the drone finds something out of the ordinary, it will approach the threat and investigate further. With the use of AI technology such as facial recognition and deep-learning identification methods, it records and documents what it sees and reports it back to a main surveillance room, where the only human oversight is required. Azur Drones Skeyetech system is the first autonomous drone in Europe approved by the authorities.
​
While Azur Drones is focused on the security of industrial properties, a company called Sunflower Labs is developing drones that securitize the common household. Their product delivers a three-part system, including garden lights called “sunflowers” that serve as activity sensors, their unique autonomous drone called the “bee”, and the “hive” - a weatherproof drone-charging station with AI-powered computing. The drones feature next-gen flight safety technology, such as ultrasonic object-avoidance sensors and infra-red lading cameras. In addition to this, the drones are geofenced to the customer's property, ensuring it never leaves the perimeter. Similar to Azur Drones’ system, the sunflowers catch anomalies in movements and triggers the bee to investigate. It then flies around the perimeter of the property, documenting and reporting any found information to the owners mobile device.
Skeyetech Drone
Azur Drones Landing Station
Google's new $1.1 Billion Data Center in the Netherlands
​
Advertisement and Marketing
​
Artificially intelligent drones even have the ability to change the advertisement and marketing industry. Because these drones can be autonomously programmed to fly in certain patterns, when used in large quantities and equipped with bright LED lights, moving images can be created in the sky above. Cities in Asia have already begun conducting these light shows for experimentation and special events.
​
At the moment, these drones are only being used to demonstrate their capabilities. However, it's not difficult to imagine the possibility of a massive coca-cola advertisement hovering over Manhattan.
​
Transportation of People
​
I must admit, using drones to transport people is highly unlikely for the near-term future, and I wouldn't include it to be "technology of tomorrow." While sizable investments have been made into the research and development of these vehicles, so much more testing and development needs to be done until they can be relied upon to the same extent as our automobiles. After all, there's much more responsibility with ensuring the safety of a passenger than the safety of our food. Because of this reason I will avoid sharing my excitement about it. However, once the day comes, the ability to travel via drone will completely reshape the entire real estate market as we know it. Daily commutes may no longer be such a hassle, as flying above traffic could arrive a passenger at their destination in minutes. Proximity to metros, amenities, or workplaces would no longer be of such importance, allowing families to live in more remote locations and have even quicker access to where they would want to go than they would previously. At the moment, home prices are heavily measured based on this proximity, and stripping this pricing factor away would revalue the market for homes in almost every location.
​
​
World's First Drone Taxi Completes Test Flight in Dubai
Militarization
Intelligent Drones in warefare
.jpg)
General Atomics MQ-9 Reaper Drone
~ $16,900,000 Production Price
​
Unmanned aerial vehicles have been utilized by militaries for over a decade now. In fact, it is known that the number of countries with military drones has skyrocketed to almost 100. Originally, the most attractive upside to this technology was to pilot a weaponized aircraft from afar, reducing accountability and increasing secrecy. The downside to this, however, was the massive price tag for production, often ranging from $100,000 to hundreds of millions of dollars. These military drones have caused much controversy over the ethicality of using such powerful weapons. Yemeni Activist Farea Al-Muslimi expressed this concern in a Senate Judiciary Subcommittee Hearing in 2013, urging how such firepower used by the United States in foreign territories goes beyond the realms of what is acceptable.
​
This was seven years ago, and as advancements in technology increased, so has this concern. Additionally, governments are continuing to get as much of an advantage with military drones as possible, and according to Mark Bowden, U.S. Air Force pilots training to fly drones outnumber those training to fly piloted aircraft. The question that remains now is a matter of when these drones will become smaller, cheaper, and more intelligent. Better yet, what happens when these military drones function by the use of autonomy? By incorporating A.I. technology, it’s certainly a possibility, and creative minds alike have already begun imagining a future where these devices exist for purchase. The following is a 7-minute film produced by ALTER called "Slaughterbots", demonstrating the creative mind of Stewart Sugg. This fictional video follows a presentation unveiling the release of a new form of A.I. weaponry with drones set in a dystopian future. The story takes a dark turn once the true possibility of such A.I. autonomy in drones is actually used.
RQ-4 Global Hawk Size Comparison
DJI's Mavic Mini Size Comparison
​
By using advanced artificial intelligence technology, all these drones need is a way to profile a target. It could characterize by age, gender, or even ethnicity, and would autonomously decide who lives and who dies. Such small devices would be cheap to produce as well, and as witnessed in the video, releasing hundreds of these synchronized devices seems to cost far less than what a non-autonomous military drone costs today. "Enough to kill an entire city" the CEO explains. "Nuclear is obsolete. Take out your entire enemy virtually risk free."
​
While we haven't yet reached a level of A.I. or drone technology where this horrifying depiction could actually take place tomorrow, the terrifying reality is that we're certainly getting close. In June of 2019, a company called Xinhuas showcased their newest drone technology at the Shenzhen International UAV Expo. The drones fly autonomously and are fitted with grenade launchers, machine guns, and missiles. Similar to "Slaughterbots", these drones utilize artificial intelligence, but only to recognize predetermined targets and the proper velocity and trajectory to launch projectiles. Nonetheless, the result is an extremely precise, accurate, and deadly device that can be mass produced, and updated.
​
While governments across the world may not know the extent to which these technologies will become weaponized, they are aware of the potential they possess. There has been multiple proactive policies set in place over the past couple years by organizations, but because these technologies advance at such a rapid pace, the ability to set guidelines is very limited. While no regulations have successfully been placed on the development and uses for artificial intelligence, efforts to regulate drones have been made. The following three organizations attempt to address these threats and concerns.
-
The Federal Aviation Administration (FAA):
A governmental body of the United States that has power to regulate all types of civil aviation across the country and in surrounding international territories. In the case of drones, the FAA has put in efforts to restrict airspace that civilians can operate these UAVs in and have figured out ways to penalize any who obstruct this rule. The FAA has listed airports and airstrips, cities, and governmental areas as no-fly zones. Any civilian who goes against these rules is going against federal law.
​
-
The Association for Unmanned Vehicle Systems International (AUVSI):
An international nonprofit organization that is dedicated to promoting and supporting the autonomous robotics industry through education and communication. They recently developed and released a “Code of Conduct”. This code states generic rules in order to observe current privacy rights and to avoid harmful accidents. Nonetheless, it is not clear yet how these rules will be enforced.
​
-
Federal Communication Commission (FCC) and the International Bureau:
Together, these two organizations have been trying to secure GPS Protection and improve performance. This is all an effort to increase the connection between drones and operators and to enable devices to better sense and avoid obstacles. Drone manufacturers such as DJI have been implementing these systems on many of their products.
​
​
​
It’s a small list, and while there may be more organizations that are making efforts to regulate these oncoming technologies, they are difficult to locate, probably due to the fact that they are still in development, along with the subject itself. Nonetheless, this short list ties back to an argument of this article - that we are entering uncharted territories when it comes to these technologies. The impacts are unrealized, but can be understood, expected, and somewhat calculated when observed through an interdisciplinary perspective.​
​
Further advancements in the fields of drone and artificial intelligence technology are being made every day by researchers, manufacturers, and developers around the world. Similar to how the introduction of the smartphone shaped the world we live in today, the applications of artificial intelligence and drones will do the same to our future. I am truly excited for the promising potential they hold and the ways they will evolve our society to further levels of productivity, accuracy, and efficiency. Nonetheless, with such powerful technology comes great responsibility, and the serious possibility of weaponization is frightening. While such an event could be catastrophic, the progress in these fields is unstoppable. This makes it greatly important that governments across nations not only acknowledge these possibilities, but also create appropriate countermeasures to combat them. The existing legislation is inadequate, and global organizations need to begin making decisions at a more rapid pace than the rapid development of these technologies.
