Over the past several years, artificial intelligence has become a larger and larger part of everyday life. Once only fodder for science fiction books, AI is now used in industries ranging from marketing to healthcare. By 2025, the global market for AI software alone is expected to grow to $126 billion.
Even though AI is quite common now, it is still a technology in its infancy. In the coming years and decades, AI and other technologies will revolutionize the world, opening up vast opportunities for economic and social development. Here’s what you need to know about how artificial intelligence will change the world.
First, Let’s Explain Artificial Intelligence
Although most people have at least a general understanding of the term, modern artificial intelligence is a broad and complex field. Before jumping into what AI can accomplish in the near future, let’s step back and take a look at what artificial intelligence really is and how it functions.
What Separates AI From Conventional Computer Programming?
To start, let’s consider how artificial intelligence differs from the programming methods used in traditional computer systems. In a conventional program, a programmer writes a set of instructions that is used by the system to process inputs. While the inputs may change, the algorithm used to process the input remains the same and is used to produce some output.
AI, on the other hand, flips this concept on its head. Rather than turning an input into an output using a set of predetermined instructions, an AI program “learns” from the inputs it is given. In other words, rather than a human programmer creating an algorithm for handling inputs, the computer develops and then continually improves its own algorithm. This allows the program to gradually get better at a given task as it is trained with large data sets and provided feedback on the accuracy of its results. Rather than using a set of rules to turn an input into an output, AI is trained to “learn” the rules of a given task as it goes.
The Types of AI
Broadly speaking, AI can be broken down into two basic categories. The first, known as narrow artificial intelligence, refers to the ability to perform a narrowly defined task or set of tasks. A narrow AI system can be used to play a game, recommend products based on patterns in customer data or manage electrical grids in accordance with power consumption. It should be noted that a task doesn’t have to be simple in order to be addressed by a narrow AI system. An artificial intelligence program used to control a self-driving car, for example, is still considered an example of narrow AI, despite the relatively complex nature of driving.
The second type of AI is referred to as general AI. Unlike a narrow AI, a general AI system would be able to deal with a range of different and unrelated tasks by learning from and applying previous experience. In this sense, a general AI would be similar to a human being. Needless to say, general AI is a vastly more advanced and technologically demanding application of artificial intelligence than narrow AI.
An extension of general AI is the concept of artificial superintelligence. Hypothetically, an AI program designed to handle general tasks and given access to sufficiently large memory and data processing resources could eventually surpass humans in almost any task. Although intriguing, artificial superintelligence is still in the realm of science fiction.
Understanding Machine Learning and Deep Learning
In addition to the basic types of AI discussed above, there are multiple approaches to implementing artificial intelligence. The two most important AI methods are known as machine learning (ML) and deep learning (DL).
Machine learning is a method by which computers are able to “learn” from data provided to them. ML systems train algorithms to properly identify patterns or other characteristics of inputs. As time goes on, the algorithm will get progressively better at correctly executing the task it is being trained to perform. ML is widely used in predictive analysis, email filtering, online fraud detection, customer support chatbots and a range of other useful tasks. In order for the training to work, however, humans generally must select the properties of the data the algorithm is being trained to identify.
Deep learning, a more advanced version of machine learning, takes this concept to the next level. Deep learning programs attempt to mimic the workings of the human brain by automatically filtering and classifying data. Unlike normal ML, a DL algorithm can seek out and learn from patterns in data even without a human telling it what to look for. Due to their ability to find patterns on their own, DL systems are typically used in more advanced applications, such as computer vision and autonomous vehicles. Despite being more advanced and more powerful, deep learning does have certain limitations. As a rule, DL algorithms require larger training data sets and more processing power, making them more costly to develop and train than ordinary ML algorithms.
The Current State of AI Technology
In order to understand where artificial intelligence is going, it’s first important to know where it currently stands and how it got there. For context, let’s take a look at the history of artificial intelligence and the status of the technology as it exists today.
A Brief History of AI
Although there were earlier precedents, the term “artificial intelligence” was used for the first time at an academic conference in 1956. That conference and its attendees set off a wave of interest in the concept of artificial intelligence, but progress in the field proved extremely slow. By the mid-1970s, the initial surge of enthusiasm had largely faded, resulting in what has come to be known as the AI winter. This period of stagnation would last until the beginning of the 1980s.
As computing power continued to improve, though, governments and businesses took a renewed interest in AI. This period would bring about considerable advancements in the field, including the development of early deep learning. This was also the era that saw the rise of expert systems, computer algorithms designed to mimic human decision making by checking inputs against an existing body of knowledge. Though primitive by today’s standards, expert systems represented an early form of AI that was useful in real-world business applications.
By the 1990s and the early 2000s, AI began to approach its current state. Early AI chatbots, computer vision programs and other applications that are relatively common today were developed during this period. At the same time, computing power continued to grow, allowing AI algorithms to become more useful and bringing about increased use of machine learning.
Where Is AI Today?
Today, cutting-edge AI programs predominantly use a form of machine learning known as an artificial neural network (ANN). ANNs model the workings of neurons in the human brain, allowing machines to “think” more like humans. ANNs allow AIs to approach more complex tasks, making them suitable for rapidly training deep learning algorithms.
In terms of applications, AI use has expanded greatly in recent years. Today, you’ll find AI programs used to do everything from curating social media feeds to improving crop yields in agriculture. Although much attention is given to the technology’s use in self-driving cars and autonomous drones, the reality is that AI systems are practically everywhere, and most people interact with them in some capacity on a day-to-day basis.
Despite its proliferation and wide range of uses, artificial intelligence is still very much an emerging technological field. Applications that are commonplace today would have been little more than academic projects 10 or 20 years ago. It’s also important to note that all modern AI is categorized as narrow artificial intelligence. Even though it can be quite efficient at certain tasks, AI is still well behind humans in terms of general intelligence. On average, most experts believe that it will be at least 2060 before general artificial intelligence is achieved.
Which Companies Are in the Lead on AI Today?
While there are any number of small startups attempting to leverage AI technology, the field is largely dominated by a handful of large, successful companies with the resources and expertise to use the technology to its fullest. Below, you’ll find a rundown of some of the leading companies in the field of artificial intelligence today.
As usual in the tech world, Google is at the cutting edge of the AI movement. Google uses various forms of AI, especially deep learning, to power many of its consumer-facing services. Voice recognition in Google Assistant, image recognition and automatic video analysis are just a few examples of how the company has implemented the technology on a practical level.
While Google is certainly a leader in implementing AI, the company’s most important contribution to the field has been the creation of an open-source machine learning library called TensorFlow. By allowing anyone with an interest in AI to develop new tools using TensorFlow, Google has massively democratized the field. As the author of the main open-source tool used to develop AI programs today, Google is arguably the global leader in the field of artificial intelligence.
While Google certainly leads in terms of AI development, Amazon has leveraged the technology to optimize its business functions and shown off the pragmatic results AI can produce. In addition to its famous AI-powered product recommendation engine, Amazon has developed the Alexa personal assistant on the customer-facing side of its business. The real power of AI at Amazon, though, is behind the scenes, where machine learning has allowed the company to massively improve its warehouse operations. Similar systems are also used to optimize the company’s delivery processes, ensuring that the countless packages handled by Amazon each day reach their destinations as quickly as possible.
Even though it isn’t as high-profile as Google or Amazon, the Bitcoin mining chip producer Bitmain has also made some important contributions to the world of AI. In April of 2020, the company revealed that it had successfully developed an AI image recognition software meant to identify rare, endangered birds for conservation purposes. Bitmain is also using its AI-capable chip hardware to support development of projects ranging from smart cities to enhanced facial recognition systems.
As you might expect, the electric automaker Tesla has leaned into AI as a key component of its business as it pushes to develop fully autonomous commercial vehicles. In fact, the company has been preparing for its AI driving system since day one by equipping every vehicle it sells with the hardware required for self-driving. As new autonomous driving features are developed, Tesla can simply send them in the form of a software update.
In 2019, Tesla even went so far as to acquire AI startup DeepScale to support development of its Autopilot system. This decision demonstrated just how important the company believes AI technology is to its future, as well as how much value startups focused on artificial intelligence technology can command in the marketplace.
Beyond its self-driving system, Tesla is also planning to deploy artificial intelligence as part of its virtual power plant project. This green energy project, meant to supplement or replace traditional power plants with home solar hardware, will use AI analytics to predict usage and demand, allowing the VPP grid to make adjustments as needed during high-demand hours.
How AI Will Fundamentally Change the World
Artificial intelligence is unlike any other tool ever developed by humanity. By allowing computers to learn and think on their own, AI systems have the potential to aid human workers in even the most complex tasks. By combining human intellect with increasingly powerful AI, the technology is poised to fundamentally change almost every area of modern life. Although exploring every single use of AI in a single article would be impossible, you’ll find some of the most important areas artificial intelligence is set to revolutionize listed below.
Healthcare and Medicine
The medical field is arguably one of the areas in which AI could have the greatest degree of impact. AI’s applications in medical settings begin with diagnostics. By scanning medical data for telltale signs of disease, diagnostic algorithms have been shown to accurately diagnose conditions earlier than human medical professionals. Impressively, this technology can work even without the collaboration of a trained doctor. In 2018, the FDA approved the use of an AI diagnostic tool for detecting diabetic retinopathy using scans of a patient’s eyes. This system operates independently of a specialized diagnostician, requiring only a lower-skilled worker to take the scans. Such systems could one day allow for faster, more accurate diagnoses in a primary healthcare setting.
The role of AI in medicine doesn’t end with diagnostics, though. The technology can also be applied to make the drug discovery process both faster and less expensive. Using databases of clinical trials and academic research papers, AIs can quickly identify candidate compounds that are known to interact with the pathology of a given disease. By comparing tissue samples from patients with and without a specific disease, AI drug discovery systems can also uncover new information about how that disease progresses in the human body. These insights can then drive recommendations of chemical compounds that could prove effective, even if they have not previously been associated with the condition in question.
Even in the high-stakes area of surgery, healthcare professionals and researchers are discovering enhanced roles for artificial intelligence. By learning from previous surgical plans, AIs can assist surgeons by proposing new surgical plans for similar cases. Artificial intelligence integration could also greatly improve surgical robots, allowing them to perform a wider and more complex range of tasks to more effectively assist the supervising human surgeon.
Finally, AI predictive analytics may one day be able to identify potential epidemics and aid infectious disease specialists in coordinating responses. This capability was demonstrated at the outset of the COVID-19 pandemic, when a handful of AI programs provided early warnings of a possible respiratory disease outbreak in Wuhan, China, more than a week before the WHO officially recognized the beginnings of the outbreak. With future improvements, such predictive analytics programs could provide earlier and more actionable insights into emerging healthcare threats. Similar systems may also be used to prepare hospitals to handle large numbers of patients by optimizing resource usage and allowing hospital staff to plan for high-demand scenarios in advance.
Taken as a whole, these uses show a role for AI in the medical field that stretches from research to practical patient care. Any one of these technological advances would be extremely useful on its own, but their collective development will revolutionize the ability of medical professionals to understand and address human diseases.
Business and Finance
Just as in healthcare, AI is set to change and optimize practically every part of the modern business world. The impacts of artificial intelligence in business begin with providing data-driven insights for decision making. From allocating resources in marketing campaigns to deciding which candidate to hire for a job, AI analytics can help businesses to make better, more profitable choices. These systems are particularly well-suited to the increasingly digital nature of the workplace, as modern businesses generate immense amounts of data that AI algorithms can learn from to optimize business operations.
The ability of artificial intelligence to quickly analyze the workings of complex systems also makes it perfectly suited to the field of supply chain management. In recent decades, global supply chains have grown longer and more complex than ever before. Using AI, however, businesses could proactively predict demand, make warehouses more efficient and optimize delivery routes to ensure that the supply chain runs as smoothly and quickly as possible. According to a recent survey conducted by McKinsey and Company, 61 percent of businesses expect to see savings in supply chain management as a result of AI integration.
It’s important to note, however, that AI’s influence in the business sector doesn’t end with logistics, planning and analytics. In the coming years, artificial intelligence will likely have just as large a presence on factory floors as in corporate boardrooms. The same McKinsey survey mentioned above suggested that 64 percent of businesses expect to see savings in their manufacturing operations as a result of AI. In large part, these savings will be the result of predictive maintenance systems that monitor equipment performance and schedule maintenance to minimize downtime. Generative algorithms can also be used to optimize digital product designs, potentially catching design flaws early and thereby reducing the length an cost of the prototyping process.
AI is also poised to have a massive impact on the financial side of the business world. The most prominent example of this is the Fintech industry, where AI is already being leveraged to provide affordable, accessible financial services at large scales. From investing to credit decisions, AI is playing an increasingly large role in financial life. As time goes on and algorithms continue to improve, it is likely that artificial intelligence will be used to optimize lending, manage financial risks and even analyze historic stock market data to improve the allocation of resources in investment portfolios.
As with the medical field, AI seems set to take on an end-to-end role in the business world. Beginning with product design and ending with delivery logistics, artificial intelligence can optimize practically every step in the process of providing goods and services to consumers. Together, these developments will make businesses more agile, more responsive to changing consumer demands and more profitable.
As we’ve already mentioned, AI can play an important role in optimizing product designs. The technology’s abilities in the field of engineering, however, are much broader. One of the most important roles for this emerging technology is integrating datasets from multiple engineering projects. Thanks to its ability to analyze huge amounts of data, AI can glean useful insights from multiple projects at the same company and deliver those insights to engineers working on each project. This, in turn, can improve cooperation and coordination between engineers responsible for each individual part of a larger design effort.
AI can also free up engineers to perform high-leverage work by automating time-consuming tasks that normally slow down the design process. In this sense, advanced AI systems will incorporate themselves into the engineering workflow in much the same way that traditional computer aided design (CAD) programs did decades ago.
The real power of AI from an engineering perspective, however, comes from the technology’s ability to help engineers and researchers develop novel materials for specialized uses. AI can support advanced material science by modelling new materials and predicting their properties. This ability to accurately predict the properties of a new material will enable material scientists to produce new materials more quickly than conventional research methods currently allow. This, in turn, will expand the range of materials available to engineers for their projects.
Solving Societal Problems
In addition to its immense potential in the private sector, AI also has a large part to play in resolving pressing problems facing modern society. Although solutions to these problems have eluded government and academic experts for decades, AI could be a critical tool in helping humanity tackle some of its biggest challenges.
First and foremost among these, of course, is climate change. AI technology is uniquely positioned to help scientists and public policy makers understand climate change by creating more accurate models of its effects. These models can also be used to determine which actions and policies are most likely to reduce atmospheric levels of carbon dioxide, allowing governments to intervene in targeted, impactful ways. The technology could also affect carbon emissions directly by optimizing power usage and driving more rapid advances in green energy technology.
In terms of more mundane human affairs, AI could also be a critical tool in the ongoing global effort to limit public corruption. In much the same way as predictive analytics can be used to detect fraud in private sector transactions, AI systems could be used to uncover inappropriate use of resources in the public sector. It may also be possible to reduce the chance of corrupt activity occurring in the first place by automating systems that were previously dependent on humans. Such automation would limit the opportunities for corrupt activity by eliminating the potentially unreliable human element and ensuring that public funds were distributed properly.
Although its role in this area will likely be smaller, artificial intelligence may be able support efforts to increase access to affordable housing. AI systems that more accurately model risk and can make lending decisions autonomously would reduce the cost of initiating loans. By using data beyond a simple credit score to make credit decisions, algorithms can also help people who are not creditworthy in the traditional sense finance home purchases.
The Promise of General Artificial Intelligence
As revolutionary as the possibilities discussed above may seem, they are all realistically achievable with narrow artificial intelligence. If general AI is ever achieved, its capabilities could expand far beyond even what has been discussed thus far. Futurists believe that a general AI could achieve seemingly impossible goals such as ending wars or eradicating human poverty. Some thinkers are even more audacious, suggesting that the development of general artificial intelligence could be a stepping stone toward rendering humans digitally immortal by eventually allowing human consciousness to run on computer hardware.
How Does it All Add Up?
As you can see, there is practically no area of day-to-day life that AI won’t have at least some effect on as the technology improves and its use becomes more common. From work and health to solving some of the world’s largest problems, artificial intelligence will be working in tandem with human ingenuity to improve almost every aspect of our lives. As a result, the coming decades are apt to be a period of rapid change and large opportunities for societies, businesses and individuals alike to leverage the power of AI.
How AI Can Interface With Other Emerging Technologies
As clear as it is that AI holds immense promise even in its current state, its real power will be brought out by other cutting-edge technological developments. In this section, we’ll discuss AI’s relationship with the internet of things, quantum computing and 5G communications networks. While these technologies are revolutionary in their own rights, they also have integral parts to play in allowing artificial intelligence to reach its full potential.
Internet of Things (Iot)
Over the last several years, the number of internet-connected devices has increased tremendously. In addition to smartphones and computers, everyday items from thermostats to washing machines have gained internet access. Collectively, this worldwide network of devices is known as the internet of things, or IoT. As of the end of 2019, the IoT consisted of about 7.6 billion devices worldwide.
When combined with AI, these IoT devices can become far more efficient and useful than they currently are. AI algorithms can be used to improve efficiency in operations by analyzing data from workplace IoT devices. Similarly, connected devices in the workplace can interface with AI systems that can predict and mitigate risks before they have a chance to cause serious business disruptions.
AI can also be deployed to improve IoT devices themselves. AI systems integrated into drones, robots, self-driving cars and other smart devices can allow them to function independently and performs tasks that would normally require human control. Through such devices, AI can actually interact with the physical world, rather than simply providing insights based on digital data.
Quantum computing is perhaps the technology that holds the greatest promise when used in conjunction with AI. This is due to the immense computational power that allows these devices to process data more quickly than any conventional computer could ever hope to. Today, the fastest quantum computer in the world can complete a calculation in about 200 seconds that would have taken a conventional supercomputer millennia to execute. Given that this technology is still in its early stages, there is little doubt that the coming years will bring even faster and more powerful quantum computers.
Thanks to this greatly enhanced speed, quantum computing could allow AI algorithms to solve large, complex problems that would currently require prohibitive computer resources. With this technology, AI could be applied to even the most complicated challenges.
Aside from simply solving specific problems given to it, an AI system powered by a quantum computer could also find previously undiscovered patterns in even the largest data sets. These patterns could then be used to derive useful insights that wouldn’t be immediately apparent to a human, expanding the potential range of solutions generated by AI.
To fully unleash the potential of artificial intelligence working in conjunction with IoT devices, large amounts of data must be transferred quickly. This is where 5G networks come into play. With transfer rates of up to 10 gigabits per second, 5G technology can handle the immense amounts of data that IoT devices will soon be feeding to AI algorithms.
Interestingly, this relationship is expected to run both ways. While 5G networks will support AI by allowing the free flow of data, it is very likely that those networks will themselves be managed by AI systems. Using predictive analytics powered by AI, telecommunications companies could predict spikes in traffic and allocate resources accordingly. This is just one example of how AI can form symbiotic relationships with other technologies as they develop together.
The Downsides of Artificial Intelligence Technology
For all of its promise, AI technology still also has certain downsides. While criticisms of the technology range far and wide, two of the most common problems raised about widespread use of AI are its potential to automate traditionally safe jobs and a phenomenon known as the black box problem.
The Unintended Economic Consequences of AI
Since AI first became a reality, critics have argued that it will replace humans and cause massive job losses. While the effect of AI on employment is likely overblown, there is some truth to the idea that the technology could disrupt the labor market and temporarily displace workers in certain roles.
To understand the full impact of these displacements, it’s first important to look at how broad AI automation can be. Traditionally, automation has had its largest impact in the manufacturing sector, where it has allowed blue collar workers to spend less time on simple, repetitive tasks. AI, on the other hand, has the potential to automate tasks that previously required human thought. Professionals in healthcare, law and engineering are among those whose jobs will likely be exposed to the next wave of technological changes in the workplace.
Although there is broad consensus that many jobs will face some degree of automation as a result of AI in the near future, estimates of the number of jobs that will be lost vary widely. Extreme estimates suggest that as much as 30 percent of the current global workforce could be displaced by technological changes by 2030. Other projections, however, are much more moderate.
The Black Box Problem
One of the thorniest challenges facing AI is the so-called black box problem. This phenomenon occurs when advanced AI, such as a deep learning system, is applied to complex problems. Often, the algorithm will act as a “black box,” taking inputs and generating outputs to solve problems in a way that even human AI designers can’t fully understand. In other words, even the humans responsible for creating AI systems are sometimes incapable of explaining how or why they reach the conclusions they do.
The black box problem presents several issues for the implementation of increasingly advanced artificial intelligence. Without a proper understanding of how an AI solves the problem it has been given, humans are much less likely to trust the answer the system provides. In some cases, AI systems produce nonsensical results due to a lack of relevant inputs. Thanks to the black box problem, these results can seem indistinguishable from correct answers, since there’s no good way to tell how the system reached its conclusion. As a result, the black box problem has profound implications for the accuracy of AI systems and human trust in them.
In some instances, of course, inaccurate results or lack of human trust are small problems. An AI-powered chatbot failing to provide a relevant answer to a customer question, for example, is inconvenient but relatively harmless. In more important tasks, though, humans must have a high level of trust before AI technology can realize its full potential. A consumer who believes that an autonomous car is likely to make a mistake and cause a traffic accident, for instance, likely will not choose to buy that car. Such lack of human trust can substantially hold back the adoption and development of artificial intelligence.
This problem can even result in machines developing biases that have real-world effects. AI systems used for evaluating risk levels in the legal system, for example, have famously exhibited racial bias by labeling members of minority groups as being more likely to commit crimes in the future. Although it is known that AI algorithms reach these decisions based on the data inputs they are trained with and given to evaluate, the black box problem makes it difficult to remove this bias from the system.
Can AI Overcome These Challenges?
Overall, the benefits of AI make it extremely worthwhile to devise solutions to the problems it presents. Fortunately, these challenges aren’t as insurmountable as they appear on the surface. In the case of the black box problem, greater transparency regarding the inner workings of algorithms is a likely basis for a solution. By allowing humans to more easily discern how an AI system reached a given conclusion, transparent systems could solve or at least substantially mitigate the black box problem.
As for the economic upheavals caused by AI automation, it’s important to keep in mind that new jobs will be created as old ones become obsolete. This concept, known formally in economics as creative destruction, has held true in other periods of massive technological change. In fact, there are some estimates which suggest that AI will be a net creator of jobs. A 2018 World Economic Forum report found that AI technology would create 133 million jobs by 2022 and displace only 75 million existing workers. Assuming these projections are correct, the net effect of AI automation would be a gain of some 58 million jobs. If job training is made available to help displaced workers find new roles in the modern economy, artificial intelligence could unlock new, better-paying jobs for millions of people worldwide.
The Future of AI
As you can see, the opportunities associated with artificial intelligence in the coming years are nothing short of enormous. This technology has the power to transform industries, create new ways of working and even solve some of the largest problems faced by our society. While no one can accurately predict every use that AI will have months or years from now, it is extremely clear that AI will be one of the driving forces of the global economy for the foreseeable future.
by Ben Carmitchel
Use my referral link to receive 1,000 free Supercharging miles with the purchase and delivery of a new Tesla car: http://ts.la/benjamin6850.