What Exactly is Artificial Intelligence Anyway?
In a previous post I wrote about how, throughout history, technology has focused primarily on solving all sorts of complex problems affecting human beings and their environment. Despite the positive results achieved, this approach has a downside; individuals need to know how to use technology and have to actually use it to produce an outcome. Most of today's machines, vehicles, computers, are examples of such technology. In the future, however, technology will focus on humanizing its use, smoothing out our interaction with it. Thus, individuals will no longer need to know how to use technology; technology will tell them how to use it. Additionally, technology will free up a huge amount of time, since humans will no longer need to be involved physically in its use. Driverless cars are a case in point.
And this is when artificial intelligence (AI) comes to life. In its most simple definition, artificial intelligence may be considered an interface (with most of the attributes that define a human being, such as language, knowledge representation, reasoning, self-awareness, learning, adaptation, sensing) between human beings and knowledge and technology. Furthermore, since human beings use technology to explore, analyze and deal with reality, AI may become the interface between humans and reality. AI will provide human-like solutions to human-like problems in a human-like (though more "efficient" or "purely rational") way.
AI, which could be regarded as a technological paradigm instead of just as a tehnology, may be the long awaited interface layer below which innovation and breakthroughs will continue to shake our society and above which human beings will be able to acquire, understand and share new knowledge and access and use new technology. Considering the dramatic increase in life expectancy during the last century (30 years on average) and the potential future increases combined with the accelerated pace of technological and scientific advance and the exponential growth of knowledge it becomes clear that human beings alone will not be able to keep up with the tide of changes and innovations they will witness during their lives. AI will be our perpetual gateway to knowledge and technology. And the ultimate learning experience, since it will be able to tailor any teaching strategy to our specific needs and personal characteristics. AI will contribute to improving everyone's mental skills and cognitive abilities, sharing the benefits of humankind's advancement, fostering an enhanced cognition (not only through improved thought processes but also superior sensing performance and increased data processing capacity, a combination of software, hardware and connectivity, a blend of augmented and virtual reality).
As a digression, let me add that human language was the first interface layer. It allowed humans to conceptualize and reason about reality (outer world) and express their feelings and emotions (inner world). There were no longer just humans and reality, there was language in between. And language somehow shaped our relation with and experience of reality. By the way, that was the moment of singularity for the human being, when we started generating and manipulating concepts and abstractions, paving the way for science, technology, social and economic institutions, and many other human creations like art. There won't be any true singularity in the future, just incremental improvements, even if extraordinary. AI may be regarded as a new kind of language that not only helps conceptualize and reason but also learn, adapt and act. Actually, AI machine translation tools (like Google's) are thought to be developing a kind of "interlingua", an interface language between translated languages.
Of the many elements AI is characterized to have, natural language processing (to interact effectively with humans), knowledge representation (to store and manipulate acquired knowledge) and reasoning (to analyze and infer, draw conclusions and make decisions, plan and act) will be paramount in its role as interface. Other necessary features like sensing (computer vision and others), interacting with the environment (robotics, crucial to embodied cognition), machine learning (to actually learn and adapt to new conditions) and many more, will be part of the applications and technologies developed below the interface layer, and will be accessed on an 'ad hoc' basis (mainly through the Internet of Everything, see next post).
As an interface layer, AI will not only allow us to access knowledge and use technology (kind of weak or narrow AI), but will also gain insight into our thought processes and behavior (the inner workings of the brain), learning from us and becoming more able to interact with us and offer adequate solutions to our needs, even anticipating them (strong or general AI, which could be regarded as an evolutionary result of human-machine interaction).
Additionally, AI capabilities (as an interface layer) will be subject to human guidance, what some people have come to call "human in the loop", which not only will improve results (since perfect rationality may not be possible in complex environments) but also bring control over actions taken (active learning, reinforcement learning and "emotional/moral" learning), ensuring its beneficial use.
AI might be external to humans, embedded into standard (machine, computer, smartphone) or anthropomorphic (robot) devices, or most probably internal, integrated into a human body or brain (Elon Musk's neural lace concept). The main point, though, is that AI is an interface, not a standalone reality. Some people may claim that the only way for humans to not be overtaken by intelligent machines is to become machines themselves. In my opinion, AI might be one of the most extraordinary human achievements ever and it just wouldn't make sense to not integrate it into human nature. Why would we want to build intelligent robots (smarter than us) instead of enhancing our own mental abilities and be able to monitor thousands or millions of robots?
Once more, AI is not about creating something very similar to us that will eventually replace us but about helping us solve our problems (acquisition, understanding and sharing of knowledge and access and use of technology) and deal with our challenges (health, education, unemployment, inequality, violence, overpopulation, climate change, biodiversity, water, energy, etc.) and, among them, reach a deeper understanding of how the brain works and what human beings actually are or might be. Hence, we need not worry about AI since it should always be human and controlled by humans, most probably integrated into our own bodies. Well, we just have to worry about humans.