Why not artificial consciousness?

VICTOR KANNAN explores the debate around whether AI has consciousness. He comes up with a very compelling logic for how it can easily be solved.

There is a growing debate about whether AI has consciousness.

I find this question amusing. The seriousness with which we debate this question seems to be misplaced. After all, we are not asking if an intelligent being is devoid of consciousness. If we are not asking that, why then are we asking if programs that are powered by artificial intelligence have a human type of consciousness.

We seem to have understood intuitively what artificial intelligence is and what consciousness is. Because the responses we get from our prompts to Gemini or Chat GPT seem to be personal and contextual, we are surprised, and we ask if AI is conscious. 

What is intelligence?

According to the Oxford dictionary, “intelligence” is the ability to acquire and apply knowledge and skills. When this is done by programming a computer, it is artificial intelligence as opposed to natural intelligence, which is innate, such as found in living beings like plants, animals and humans. They use innate intelligence to respond to situations, while machines are programmed to respond. It is a different discussion to ask if we are not programmed or influenced by our past memories and present socio-economic conditions. This takes us to behavioral science and psychology.  Let us, however, stay focused on this question of AI having consciousness. 

What is consciousness?

According to the Cambridge dictionary, “consciousness” is “the state of being awake, thinking, and knowing what is happening around you.” Consciousness is the degree of awareness and unawareness, thinking and not thinking, knowing and not knowing. Somehow this definition is short on the sleep and dream states of consciousness. 

What does AI do when it is “sleeping”?

We might say AI is learning while “sleeping” based on the previous set of interactions. This is machine learning. We might say this is similar to how the brain functions while a person is sleeping. Just as the brain consolidates and stores information obtained during the day, AI machines are continually indexing the knowledge obtained during the period of interaction for future use. This sounds like a valid argument. 

Repetition and creativity

The fundamental difference is the repetition of a machine versus the creativity of innate consciousness. We might ask the same question to AI repeatedly and it would keep on producing slightly different answers; hence, we can claim it to be creative or imaginative. The random number generator in a gaming device appears to produce unpredictable outcomes. Those who know the rule that the house always wins, know 
the sum of all outcomes is profitable for the house and not the sum of all players.

Many lose so a few can win. The predictable outcome is that the house always wins, though individual outcomes are random. The fact that the house always wins means the aggregate outcomes are pre-programmed.

Pre-programmed and uniqueness Similarly, AI machines are pre-programmed. They have artificial intelligence but not the consciousness to create the intelligence to begin with. Machine learning algorithms do not make them originally creative and unique, while human consciousness does. Consciousness is a composite of the ability to think, process, decide, and act in a unique manner based on circumstances and past learning. It all seems programmable. But two people respond totally differently to the same set of circumstances and this highlights the uniqueness in behavior.

Knowing and how we know are aspects of consciousness. Intelligence is a means to know, and hence is part of consciousness. Knowing includes feeling, and feeling instigates thinking. Thinking results in understanding, and that helps in choices and decision making. All these are aspects of consciousness. Hence AI is an extension of human consciousness; it is not that AI has any consciousness of its own or even the possibility of having its own consciousness.

Humans can behave in a mechanical way, but machines cannot behave like humans. Here lies the confusion and pathway for clarity! The best we can do is to define the existence of artificial consciousness as an explanation for the learning aspects of Artificial Intelligence engines. If we call it Artificial Consciousness would the debate end?