The strong artificial intelligence would be in the image of the man with a superior cognitive intelligence. But is strong artificial intelligence possible? Definition and examples.
What is strong artificial intelligence?
According to the theory of strong artificial intelligence, AI could develop an autonomous consciousness, sensitivity and will, on the model of human beings.
The goal of strong AI is to create a complex intelligence designed to address the full range of problems that humans are capable of grasping, as the human brain can naturally do.
Is strong artificial intelligence possible?
Several aspects of intelligence belong to the field of strong intelligence:
- cognitive intelligence (or cognitive computing),
- emotional intelligence,
- social intelligence,
- psychomotor intelligence.
Today, the concept of strong AI is theoretical rather than practical. Most of the programs based on artificial intelligence appeal mainly to cognitive and cognitivist abilities:
- rationality,
- organization,
- planning,
- learning,
- autonomy,
- problem solving.
Examples of strong artificial intelligence are still in the realm of science fiction. There are machines integrating AI, with great performances, in particular cognitive ones, but these machines are not provided with any autonomous consciousness.
For some, the exceptional performance of deep learning, based on artificial neural networks with a complex architecture, augurs well for strong AI. A concrete example of strong artificial intelligence, which is still highly perfectible, is the autonomous car. In its most accomplished version, it can make a decision (brake, stop, turn) and avoid some incidents, but not all.
Today, the autonomous car can assist the human being, and even, for some models, ensure 98% of the driving load. But the other 2% must still be managed by human intervention. The autonomous car has neither consciousness nor sensitivity and, faced with particularly complex situations, is not able to react.
What is the greatest artificial intelligence?
The largest machine learning models are the large language models (LLMs). One of the most famous of them is GPT-3 (for Generative Pre-trained Transformer 3). Unveiled by the company OpenAI in 2020, it is a giant neural network for automatic language processing with no less than 175 billion parameters. It is based on a Transformer architecture.
What is the difference between strong and narrow artificial intelligence?
Weak artificial intelligence uses AI for specific tasks. It can perform calculations, process large amounts of data, solve specific problems and learn in an automated way. But it has no consciousness of its own and acts according to the way it has been programmed by humans.
Strong AI is the AI vision of tomorrow, it aims to create autonomous machines with consciousness. Today, the AI integrated into applications and machines belongs to the weak AI. Strong AI, on the other hand, is the challenge and certainly the most ambitious project in AI.