Artificial Intelligence is in the headlines. You can't escape it. Some fear its job-destroying impact, others its threat to human self-worth. The immediate future of AI is a lot more complicated -- and promising.
People are confused. AI is either considered to be a giant leap towards modernization or a catalyst for massive unemployment. For the foreseeable future, it will be neither. AI will gradually affect everything we do; it will create more jobs than it destroys, and it will improve everything from customer experience to patient care. It all depends what we do with it. In other words, the future of AI lies in the hands of those controlling the characteristics of this revolutionary tool.
"A" stands for "augmentation
AI is advancing in leaps and bounds. Today we can imagine applications that seemed preposterous only a few years ago. From a computer system playing Go to driverless cars, from marketing tools to AI-directed healthcare solutions, the possibilities seem endless. The one thing that is certain is that for years to come, AI will augment humans, improve our capabilities, and help us become better at every tasks. AI will take over repetitive, data-driven tasks, and leave us more time to focus on creative, complex ones. In short, AI will augment humans, not replace them.
The superiority of machines over human brains must be understood in terms of speed and scale of computation. Anything that requires vast amounts of data and complex calculations we should leave to the machines. That includes data collection, analysis, and prediction, but not the actual result or its implementation. Humans are still better than machines at judgment and making sure that the AI's goals are aligned with our human goals. So while AI can tell you what happens when in a city, it won't be enacting traffic laws anytime soon.
Considering the high level of analytical skills AI professes, the technology will primarily be used to expand human capabilities, to optimize the use of resources, and to enhance productivity, rather than replace human actors. A few of the areas which will immediately benefit from AI are listed below.
Daily Life
When you think of how AI can assist and automate our everyday lives, just think of the voice assistants in your mobile phone, or the friendly shopping assistant in the Amazon Echo. Alexa can understand and react to commands, like setting a timer, playing music, order washing powder, or answer general questions based on web searches. It's not much of an "AI" yet, but it will quickly improve.
The more you use Alexa, the more it will adjust to your speech patterns and your tone. All the data it collects are stored in the cloud, and the more you use it, the better it will get. As millions of users provide more and more data, machine learning algorithms will ensure that Alexa grows up very quickly. For now, it won't replace genuine human advice, because ultimately all it does it recognize questions and look up the answer in a database.
The Internet of Things
Much more than voice assistants, smart sensors and other Internet-connected devices will augment these primitive AI interfaces and make everyday living easier for humans. Smart sensors take input from the environment and then follow built-in resources to carry out predefined functions. They are meant to collect vast amounts of data much more efficiently, and are already an integral part of many companies' approaches to AI. Telling Alexa to lower the ambient temperature, for example, involves measuring the actual temperature and then adjusting a thermostat. As IoT devices become ubiquitous, AI will become more useful in practice.
Manufacturing
Manufacturing businesses can benefit from smart devices that detect and regulate the usage of inputs, production parameters, or the quality of goods manufactured. With the help of collaborative robots that take over tasks that are normally considered repetitive, unsafe, and difficult, AI will be a major part of any manufacturing process. Humans will assure smooth operations, while robots, smart devices, machine learning and AI will ensure maximum efficiency.
This integration of new technologies is already being implemented by leading manufacturing everywhere; adoption will accelerate quickly as improved efficiency means higher margins and capacity.
The same technologies will be deployed in every aspect of the enterprise. Smart systems will order their own raw materials; smart sensors will track quality and notify suppliers upstream; quality and performance feedback from downstream players will improve the manufacturing process. Integration with order processing and accounting will assure that companies get better and better at their tasks.
As data usage increases, employment opportunities will multiply too. The early robotic systems won't replace workers, but rather allow them to focus on the creative solutions and implementation tasks that will always need to be the forte of human workers. AI will make mission-critical and dangerous tasks safer, faster, and more reliable.
Transportation and Logistics
The race to produce fully automated vehicles is on, but long before we get there, we will use AI to improve incrementally various stages of the transportation and logistics process. That means moving parts in factories and warehouses, reducing haulage times on long-distance operations, loading, sorting and packing, and improving fuel efficiency.
Again, humans will be play an important role throughout this process in making sure all the new technologies works together to achieve the desired outcomes.
Healthcare
Machines with inherent abilities and knowledge can help medical professionals and doctors come up with wider, more accurate, and personalized diagnosis techniques. Yes, IBM Watson can analyze more chest x-rays in shorter times than radiologists, but the optimal treatment and effective cure of diseases will, for many decades to come, involve human doctors, human experiences, and the human touch.
If anything, AI will increase employment opportunities for medical professionals, as we learn to diagnose and prevent many conditions will now lead to unexpected and unfavorable outcomes. Fewer doctors will be busy cutting out tumors, yet many more of them will be required to interpret AI-based test results, recalibrate AI-based diagnostics, refine treatment regimes, and offer counseling. If you consider that most medical staff spend over 60% of their time doing repetitive tasks like documentation, blood collection, or data analysis and comparison, AI will be a boon for the medical profession that will help to improve the experience and quality of life of all patients.
Agriculture
Our agriculture is primitive and reliant on hard-to-control factors like weather, soil conditions, availability of sunshine, scarcity of nutrients, and many other random ingredients.
AI-based, high-tech agriculture solutions, comprising, sensors, robots, data analysis, machine learning, irrigation systems, pest warning systems, fertilization and crop optimization methods will greatly increase yields and the quality of products.
These practices will help reduce the burden from humans and will also help increase agricultural productivity.
Smart retailing
Alexa & Co., combined with sensors in your fridge, can already tell you when you are running low on supplies, and where you can order them cheap and fast. Once you confirm your order, she will let you know when you can pick them up from your local outlet, or when the delivery drone will arrive.
While AI has useful applications in many other areas as well, it is really tailor-made for the retail sector. Amazon, Apple, and Google are exceeding all consumer expectations by delivering the unexpected with the aid of AI, and many local providers around the world will follow.
The massive increase in home deliveries at no extra cost may seem ideal for large countries like the US, but even in densely populated cities there are plenty of opportunities for smart retail. In stores themselves, prediction of sales numbers (more beer on Friday, more coffee in the morning) will optimize supply chains and production cycles.
The smart assistants of the next generation will predict information based on the habits of the user, and make a smart guess as to what will be required, when it will be required, and where. 24-h convenience stores in Asia are already using AI and machine learning not just to predict the number of beef noodles sold on Tuesdays, but also how much beef and how much flower to order for the next quarter.
People Algorithms
One thing AI is still very bad at is understanding its environment. But modern machine learning has solved that problem too. AI can now adapt to the changing environment and predict how user preferences can vary. Examples of ML can be found in healthcare where patients can be differentiated on the basis of what disease they are suffering from. ML, through its advancements in judging the environment, can predict the symptoms of a disease and form a solution based on its knowledge.
However, it is imperative that the algorithms and control of AI be democratized and provided to all stakeholders. The algorithms will work best when they are augmented and controlled by humans. After all, we understand our own goals and needs better than any machine (at least for now). By opening up the ecosystem and assuring participation by creators, producers, and consumers alike, we can assure better control and management over the algorithms that define artificial intelligence.
Beyond any doubt, AI, if properly implemented, will open doors to endless opportunities and possibilities, long before it becomes any kind of threat to humanity.