We have reached a stage in technology where it can either be a blessing or potentially destructive. The pace of change in the world of technology is incredibly fast. Things are evolving so rapidly that what we consider the most advanced technology today can become obsolete in just a few days.
Computers or any machines are essentially empty boxes. They require interaction to be put to work. Using programming languages, we give instructions to computers or machines. They perform tasks based on those instructions.
Whatever instruction we give to the machine, it will work accordingly. It cannot do anything beyond that unless its intelligence is enhanced. We call this enhanced intelligence artificial intelligence or synthetic intelligence. If we think about a robot, its intelligence is artificial intelligence.
Artificial intelligence refers to the intelligence exhibited by machines or robots that goes beyond their inherent capabilities. It requires some level of cognitive enhancement. If we consider the intelligence of a robot, it is artificial intelligence.
What is intelligence?
Intelligence is the ability to acquire knowledge and apply it. Regular programs cannot acquire knowledge. However, there are machines or programs that are designed in such a way that they can learn on their own. We call them intelligent programs or intelligent machines. For example, the Google search program is an intelligent program. When we search for something, it takes into account our previous search history, age, location, and other factors to show us search results. Similarly, the current chatbot is an intelligent program. MIZAR AI is an intelligent program. The goal of artificial intelligence is to provide computers or machines with human-like knowledge and the ability to think like humans.
Artificial intelligence
Artificial intelligence or machine intelligence refers to the ability of a machine or program to make decisions on its own. If a machine or program can only perform tasks based on given instructions, we cannot consider it an intelligent program. It doesn’t matter how large or complex the program is. For example, various types of robots or machines are used in call centers today. Those robots or machines can perform specific tasks for which they are programmed. However, these machines are not intelligent machines. On the other hand, there are various types of vacuum cleaners available for cleaning household dirt. They can make decisions based on whether there is dirt on the floor or not. Therefore, we can consider these vacuum cleaners as intelligent machines.
Introduction to Artificial Intelligence
John McCarthy is considered the father of artificial intelligence or machine intelligence. In 1956, at the Dartmouth Conference, McCarthy posed the question of what artificial intelligence is and how a machine can learn and demonstrate reasoning like a human. From 1956 to 1974, there was significant research in the field of artificial intelligence. Although there were theoretical advancements during that time, practical applications were limited.
The resurgence of research in artificial intelligence began in 1994. Alongside this, computer capabilities were also improving rapidly. Effective results started to emerge in this field when graphics processing units (GPUs) became readily available. Typically, computers primarily used central processing units (CPUs) that could process information one task at a time. GPUs, on the other hand, work differently. They can process multiple sets of data simultaneously.
A machine can only speak or perform tasks like a human when it acquires a vast amount of knowledge, just as a child starts with little to no knowledge and gradually learns over time. Eventually, the accumulated knowledge forms the basis for intelligent decision-making. Computers are similar in this regard. If they lack sufficient information, how can they learn? Without the ability to learn, they cannot provide any meaningful answers. In the present day, the internet contains a wealth of information. Alongside GPUs, this abundance of data has led to the creation of numerous intelligent programs.
According to various experts, artificial intelligence or machine intelligence can be categorized into three types.
ANI or Artificial Narrow Intelligence: ANI, or Artificial Narrow Intelligence, refers to a specific area of expertise in which a machine excels. For example, if a machine is designed to play chess, it will only excel at playing chess. Even though transitioning from playing chess to playing a game like Ludo is relatively straightforward for humans, the same cannot be said for the machine. If asked to play Ludo instead of chess, the machine programmed for chess will not be able to adapt and perform well. ANI is also referred to as Weak AI. We can consider it as the first step towards artificial intelligence, specifically in a narrow domain.
AGI or Artificial General Intelligence: The second step in artificial intelligence is AGI, which stands for Artificial General Intelligence. It is also known as Strong AI or Human-Level AI. In this step, computers will be capable of thinking like humans, planning like humans, solving problems, and adapting to new environments by observing and understanding their surroundings. ChatGPT, OpenAI’s GPT-3, and other similar systems are the first steps towards Artificial General Intelligence. These systems may soon be able to think and reason much like humans.
ASI or Artificial Super Intelligence: The third step in artificial intelligence is Artificial Superintelligence, which occurs when computers become more intelligent than humans. When a computer surpasses human-level intelligence, we refer to it as Artificial Superintelligence. Researchers are already concerned about the implications of creating an Artificial Superintelligence system. They are actively studying the potential consequences and considering whether it will be beneficial or detrimental to humanity.
Artificial Intelligence Singularity: Singularity is a mathematical term that represents a point where the location or values of a function cannot be determined. In the context of artificial intelligence, singularity refers to the point in time when an intelligent machine surpasses human-level intelligence. Artificial Superintelligence (ASI) is achieved when such a machine is created. The development of Artificial Superintelligence raises concerns among researchers about its potential impact on humanity, whether it will be beneficial or detrimental.
Currently, we are still in the stage of Artificial Narrow Intelligence (ANI). All the existing intelligent machines or programs are a combination of multiple artificial narrow intelligence systems. Our smartphones already have numerous ANI programs. Some of the most successful ANI programs today include Alexa, Siri, Google Assistant, ChatGPT, and Microsoft Cortana. Many of us have interacted with ChatGPT and experienced how advanced AI has become. Examples of successful ANI applications include Google’s self-driving cars and Facebook’s AI factories. Nuclear plants also use ANI systems for operation. If multiple ANI programs can accomplish such remarkable tasks, one can only imagine the possibilities when we reach Artificial General Intelligence (AGI) or Artificial Superintelligence (ASI).
The fourth industrial revolution will revolve around artificial intelligence. We are already benefiting from artificial intelligence in various ways, although it is still in its early stages. In just a few years, its usage will expand to every domain. There might come a time when robots become household assistants, salespersons in shops, or delivery agents for parcels. These scenarios may soon become a reality. We are gradually moving towards a world where artificial intelligence will have a significant presence and play a crucial role in our daily lives.
If you want to learn about artificial intelligence:
- The Intro to Artificial Intelligence course offered by Udacity is fantastic for acquiring basic knowledge in the field. All the course material is based on the book “Artificial Intelligence: A Modern Approach,” which is taught in many universities. The book is authored by Peter Norvig, who is the Director of Research at Google. He is also an instructor for this course at Udacity. Peter Norvig has written a remarkable article called “Teach Yourself Programming in Ten Years,” which I highly recommend reading. Moreover, besides this course, Udacity offers several other courses, such as Machine Learning, that you can explore after completing the Intro to Artificial Intelligence course.
- You can also consider taking the Artificial Intelligence course offered by edX. After grasping the basics, there are numerous online resources available for further learning. It’s worth taking the initiative and getting started.
- Stanford’s AI course outline, slides available here.
- If you have a lot of patience, you can watch the video lectures created by MIT. Additionally, you can download and study the lecture slides. However, please note that you shouldn’t expect to rely solely on these resources to start learning. It’s important to have a well-rounded learning experience, which may involve exploring multiple sources, practicing hands-on, and seeking guidance from instructors or communities dedicated to the subject matter.
For entertainment as well as reading about artificial intelligence, you can watch these:
Movies:
- Ex Machina (2015)
- 2001: A Space Odyssey
- I, Robot
- Matrix series
- Chappie
- Transcendence
- The Terminator series
- Star Trek series
- Tron series
- Her (2013)
- A.I. Artificial Intelligence
TV series:
- Person of Interest
- Human
- Intelligence
- Mr Robot