How will AI change the life of the average person on Earth by 2030? 100 Years of Artificial Intelligence Research is the brainchild of Eric Horwitz, former president of the Association for the Advancement of Artificial Intelligence and Managing Director of Microsoft Research, a major Redmond lab.
Every five years, a team of experts assesses the current state of AI and its future development. The first group, composed of experts in AI, law, political science, politics and economics, was drawn up last fall and decided to devote their first report to the impact AI will have on the inhabitants of the average American city. Eight key aspects of the city must change in this way over the next fifteen years.
Transport
The speed of the transition to AI-driven transport can take the public by surprise. Self-driving cars will be widespread by 2020, and it won't be just cars - self-driving trucks, autonomous delivery drones, personal robots will also become commonplace.
UBER-style auto-as-a-service is likely to displace car owners, as public transport can be abandoned or altogether on demand. Moving will be a reason to relax or work productively, will encourage people to live further from home, at the same time, the need for parking will decrease and the face of modern cities will radically change.
Data streams from an increasing number of sensors will enable administrators to model the movements of individuals, their preferences and goals, and this can have a significant impact on the design of urban infrastructure.
But people will not be thrown out of this scheme. Algorithms that allow machines to learn from and coordinate with human input will be critical in ensuring the smooth operation of autonomous vehicles. Doing it right will be extremely important, since this is essentially the first experience of society in the physical embodiment of AI systems in the urban paradigm.
Promotional video:
Home robots
Robots that deliver parcels and clean offices will become much more common in the next 15 years. Mobile chipmakers have already squeezed the computing power of the last century's supercomputers into systems-on-a-chip, dramatically increasing internal productivity and robots.
Robots connected to the clouds will be able to exchange data, learning faster and faster. Low-cost 3D sensors like the Microsoft Kinect will accelerate the development of perceptual technology, and breakthroughs in speech recognition will improve how robots interact with humans. Robotic limbs in research labs are evolving into consumer devices today by 2025.
But the cost and complexity of reliable hardware, as well as the difficulty of incorporating perceptual algorithms in the real world, means that robots for general use are far from coming soon. Most likely, robots will be limited to narrow commercial applications for the foreseeable future.
Health
The impact of AI on health over the next 15 years will depend more on regulation than technology. The most transformative capabilities of AI in health require access to data, but the FDA has yet to find a balance between ensuring access to data and respecting its integrity. Health data collection electronics are also extremely slow to implement.
If these barriers are removed, AI can automate the long and complex process of taking patient records and studying scientific literature. This kind of digital assistant will allow doctors to focus on the human aspects of patient care, to give free rein to their intuition and experience.
At the population level, data from patient records, wearable devices, mobile apps, and personal genome sequencing can make personal medicine a reality. While fully automated radiology remains unlikely, access to large medical imaging databases will enable machine learning algorithms to be trained from scans, reducing the burden on physicians.
Smart walkers, rocking chairs, and exoskeletons can help keep older people active, while home technology can support and control them to help them stay independent. Robots can perform simple tasks in hospitals, such as delivering things to the right ward, or stitching, but for the time being, these tasks will be semi-automatic and require interaction between humans and robots.
Education
The line between classroom and one-to-one learning may be blurred by 2030. Massive open online courses will enable educated mentors and AI technologies to customize learning individually at any scale. Computer-assisted learning will not replace classroom education, but online tools will help students learn at their own pace and use their favorite methods.
AI-compatible education systems will learn the preferences of individuals, but the accumulation of this data will accelerate overall learning and the development of new tools. Online learning will expand access to education, allowing people to retrain and receive higher-level education, even in developing countries.
Sophisticated virtual reality systems will allow students to immerse themselves in historical and fictional worlds in order to explore their environment or scientific objects without going directly to their location in the real world.
Poor societies
Contrary to dark sci-fi views, AI will help improve the lives of the world's poorest by 2030. Predictive analytics will enable government agents to better allocate limited resources, predict environmental hazards or violations of rules and regulations. AI scheduling will help distribute surplus food from restaurants to food banks and shelters before they go bad.
Investment in these areas is sorely lacking. There are also concerns that machine learning could inadvertently lead to discrimination based on race or gender. But AI programs are easier to hold accountable than people, so they are more likely to help avoid discrimination.
Public safety
By 2030, cities will rely heavily on artificial intelligence technologies to detect and predict crime. Automatic processing of video surveillance and aerial footage using drones will quickly detect abnormal behavior. This will not only allow law enforcement agencies to react quickly, but also predict when and where crimes will be committed. There are fears that bias in systems and errors will lead to tracking down completely innocent people, but well-designed systems can counteract human bias and abuse of office in the police ranks.
Techniques such as speech and gait analysis can help investigators and security guards identify suspicious behavior. Unlike abuse of power among law enforcement, AI can be completely independent of human and external factors and therefore more honest.
Employment and jobs
The effects of AI will be especially noticeable in the workplace. By 2030, AI will encroach on qualified professionals: lawyers, financial consultants, radiologists. As it takes on more and more roles, AI will infiltrate a wide variety of organizations and occupy a wide variety of jobs.
And yet, AI is likely to replace tasks rather than workers in the near future, and create new jobs and markets at the same time. And while it can certainly lead to lower incomes and job prospects, increased automation will also lower the cost of goods and services, making everyone richer.
These structural shifts in the economy will require political, rather than purely economic, responses to ensure that this wealth is shared. In the short term, resources will flow into education and retraining; in the long term, larger social safety nets or radical approaches like a guaranteed basic income will be required.
Entertainment
Entertainment by 2030 will be interactive, personalized, and immeasurably more fun than it is today. Breakthroughs in sensors and hardware will drive virtual reality, tactile entertainment, and companion robots both at home and outdoors. Users will be able to interact with entertainment systems verbally, and they will express emotion, empathy and the ability to adapt to changing environmental conditions such as the time of day.
Social media already allows customization of personal entertainment channels, but data streams are not yet sorted and served exactly according to users' preferences and habits. In the future, this will be brought to an unprecedented level. There is even a fear that this will give media conglomerates control over the online people, over the experiences and ideas that they pass through themselves.
ILYA KHEL