Can AI Think?

Works of science fiction have ingrained a certain perception of artificial intelligence (AI) into our society. Many want to believe that AI is capable of thinking in the same way humans do, but this is not the case, at least not yet. What users perceive to be "thinking" on the part of AI tools is just the system running established algorithms to calculate how to react. The AI follows specific steps programmed into it to arrive at its conclusions. This process can best be explained by contrasting it with how the human mind thinks.

How Human Thinking Differs From AI "Thinking"

Humans commonly consider the act of thinking to be a single process, but it consists of two different systems.

• System 1: Automatic thinking that happens unconsciously in response to stimuli
• System 2: Slow and logical thinking that relies on rationalization to solve complicated problems

System 1 thought is used for things that we instinctively know how to do. System 2 thought occurs when we need to evaluate a situation and solve a problem, like trying to find a friend in a crowded room or locate a misplaced object. System 2 thought is a process that humans are conscious of doing.

It is the system 2 way of thinking that AI has not been able to grasp yet. As humans, we use our personal experiences to arrive at decisions when using system 2 thinking. AI, however, uses data filtered through various systems and algorithms to arrive at specific conclusions. The main difference between these two approaches is that humans may make different decisions on different days, but AI will always come to the same conclusion.

For example, imagine that the route you usually take to drive home from work is closed, so you must choose another. Humans will consider the alternatives and then make a decision based on any one of a variety of factors. Maybe you want to take a more scenic drive home, even though it will take longer than another possible route. AI, however, will always make the fastest and most efficient route the primary option.

How AI "Thinks"

So can AI think in the same way as humans? No, not yet. However, there is a philosophical argument to be made that what AI does is its own form of thinking. This topic is highly debated with no clear right or wrong answer. To come to your own conclusion, you must first understand how AI works.

AI is able to function due to several different systems and technologies, which include:

• Machine Learning: A process that allows computer programs to use algorithms and data to learn and improve over time
• Deep Learning: Algorithms layered over one another to create an artificial neural network similar to that found in the human brain
• Cognitive Computing: The computer process of simulating human thought to analyze images, speech, and text
• Natural Language Processing: The ability of a computer to understand text and spoken words similar to how a human does
• Computer Vision: A computer program's ability to analyze and understand images

Each of these systems is essential for AI to work because they enable it to use the three cognitive processes humans often perceive as thinking. These are:

1. Learning: The process of acquiring data and the rules that dictate how it is analyzed
2. Reasoning: The process AI uses to choose which algorithm to apply for a specific task
3. Self-Correction: The process of fine-tuning algorithms so AI provides the most accurate results possible

It can appear that the AI is thinking when all of these systems and processes are working together. Whether or not this actually constitutes thinking is up for debate, but the fact remains that it is fascinating and impressive, regardless of the answer.