AI: Should we talk about Computer Intelligence?
Can AI be described as intelligence without consciousness, and how does this affect our perception of its capabilities? Does AI actually learn anything, or does it simply adjust statistical models based on data? How might the increasing integration of AI into our daily lives reshape our understanding of intelligence and consciousness? Faced with the constant evolution of AI, how do we anticipate its impact on our thinking in the next 10 years?
Pierre Aumont
1/8/20242 min read
EXCITEMENT, CURIOSITY & FEAR
Until November 2022, AI was often seen in science fiction terms, where self-aware machines rival human intelligence.
Does this question make sense: "Can we talk about computer intelligence?
Let's rephrase it as: can we talk about computer consciousness? And let's not answer it! Because if we're already there somewhere in the world, this information isn't public.
All we're left with is exciting speculation and fantasies about emergence theories (roughly speaking, the fact that something highly organized and structured can emerge from highly complex systems, which in this case would be consciousness...).
Let's return to the question of computational intelligence rather than consciousness. We can't escape the need to define terms... at least intelligence. In view of all the possible definitions, I propose to start from etymology: intelligence derives from the Latin words "inter" and "legere", meaning "to bind together".
If we stick to this definition, then yes, AI is a form of artificial intelligence, but without consciousness. Indeed, it analyzes, selects and executes actions from among several alternatives, according to a set of pre-programmed rules and patterns. However, it is crucial to note that AI, in its current form, does not possess consciousness. It lacks the ability to understand or subjectively feel its actions.
Can AI be described as statistical intelligence?
Why? Because it relies heavily on statistical models to make predictions, take decisions and learn from data. This is particularly evident in the fields of machine learning and deep learning. AI algorithms analyze huge amounts of data, identify patterns and trends, and use this information to make what look to us like decisions. However, AI has no awareness of the answers it gives. It simply performs the tasks for which it has been programmed, without understanding the context or meaning of those tasks.
But then, does an AI learn anything?
If we consider that enriching its internal database is tantamount to learning, then the answer would be yes. With each interaction, the AI receives feedback from the user and adjusts its models accordingly. It "learns" from these interactions and improves its accuracy and efficiency over time. However, it's crucial to note that this learning process is different from human learning. AI doesn't learn from experience or understanding, but by adjusting its statistical models.
In conclusion, and from my point of view, artificial intelligence is a complex form of computer intelligence that analyzes, learns and makes decisions based on statistical models. However, it is not conscious and does not understand the implications or significance of its actions.
Will this still be true in 10 years' time?
