The trilemma related to the rise of artificial intelligence and the way to address it

Introduction

The influence of artificial intelligence on the present spectrum of technology is really phenomenal. We can hardly imagine a sector or an application that does not require inputs from artificial intelligence in the age of industry 4.0. Artificial intelligence has influenced various sectors ranging from retail to logistics and finance to manufacturing. The application of artificial intelligence is also significant in the domains of education and healthcare.

Although we have made large strides and progress in the domain of artificial intelligence, it is believed that we are still at the nascent stage of this long and herculean journey. We would need to take some more prospective steps in order to progress into the next stage. For instance, we would have to provide the best AI courses in the academic sector so that early training and innovation can allow us to reap the potential of human resources. The next stage would be to translate theoretical learning into practical research. This is only possible when there is a close coordination between the academic sector as well as the industrial sector.

The trilemma related to the rise of artificial intelligence

There are three important cornerstones that command the rise of artificial intelligence. It is important to understand this trilemma with the help of a timeline. During the last quarter of the 20th century, artificial intelligence was beginning to influence the development of different technologies. Needless to mention, artificial intelligence created a buzz in the business world, the academic sector as well as the research centres around the world. There were three important opinions associated with artificial intelligence.

• The first opinion was that artificial intelligence would witness a gradual rise in its application before it reaches a point of saturation.
• The second opinion was that artificial intelligence would progress by leaps and bounds and applications of artificial intelligence would continue to proliferate in the time to come.
• The third important opinion was that artificial intelligence would act as a limiting factor for growth of human intelligence as it has the potential of surpassing human Intelligence and capability

During the first decade of the 21st century, machine learning as a technology started to develop within the larger umbrella of artificial intelligence. Machine learning acted as the operating arm of artificial intelligence. With the help of machine learning, large data sets could be fed as input into the machine so that signs of progress and intelligence could be slowly inculcated into them. The growth of machine learning and the focus on interpretation of information with the help of big data popularised this technology. Supervised learning made it possible to train different kinds of models with help of labeled data.

With the help of unsupervised learning, unlabeled data sets were used to train the machines. Semi-supervised learning made use of both these techniques and reinforcement learning allowed the machine to learn from its immediate environmental settings. The goals of machine learning were taken forward with the development of deep learning technology during the second decade of the 21st century. Deep learning operated on the largest possible data sets and interpreted the meaning behind these data sets with a lot of ease. This was not possible earlier but became a reality with the help of artificial neural networks. Artificial neural networks were conceived as an analogy to the human brain and operated with the help of small functions aggregating together to perform the largest tasks. Image classification and feature extraction became possible with the help of deep learning technology.

Addressing the trilemma by categorizing artificial intelligence

We can categorize artificial intelligence into three main types taking into consideration a classical approach. The classical approach takes into account parameters like logic and goals to make the three tier classification of artificial intelligence. This classification also became possible with the goal and relevance of artificial intelligence in the present as well as the future.

• The first type of artificial intelligence is called narrow AI. This is called narrow AI because the machine is given a single specific task and the machine becomes proficient to perform that particular task over a period of time. This means that the machine is not generalized to perform other tasks and execute other applications. An example of narrow artificial intelligence can be understood from Google Translate that has been conceived to perform a specific function of language translation.
• The second important type of artificial intelligence is called artificial general intelligence. As the name indicates, artificial general intelligence is able to execute various tasks and operate on a broad spectrum of applications. The present range of artificial intelligence technology that we are experimenting with can be classified under this category. Artificial General Intelligence allows machines to understand various challenges and problems in an autonomous manner. Over a period of time, machines become proficient and do not only derive solutions to existing problems but also to newer ones. In this way, artificial General Intelligence allows a machine to perform multitasking which is a characteristic of human beings.
• The third type of artificial intelligence is called super intelligence. This is the category of artificial intelligence that has been widely debated and is regarded as a threat to human existence. It is believed that super intelligence will allow the machines to surpass human intelligence and acquire greater wisdom and problem solving capacity. It is well possible that humans would lose control of super intelligent machines because they may question the gray areas which we haven’t even thought of.

The way ahead

The need of the hour is to provide a controlled platform for development of artificial intelligence technology. The research and development in this domain needs to progress in an unstoppable manner. This is extremely crucial for deriving solutions to those problems that have created a crisis situation. For instance, the problem of climate change, disaster management and the energy crisis are the biggest challenges that we would need to overcome in a short span of time. This is not possible without the deployment of applications that are a consequence of artificial Intelligence and related technology.