Researchers from the University of Geneva succeeded in modeling an artificial neural network capable of learning tasks solely from written instructions and then communicating those tasks to another AI, which could perform the same tasks with no prior training. It marks a significant step forward in AI development, as this process mimics how humans learn and share knowledge.
The research team equipped the AI networks with natural language processing (NLP). NLP is a subfield of AI that focuses on enabling computers to understand and generate human language. By leveraging NLP, the AI networks could comprehend written instructions and perform tasks based on them.
The researchers trained the AI networks using pre-existing NLP models and neural network models inspired by the human brain. This combination allowed the AI to interpret written instructions and execute tasks with remarkable accuracy, achieving an average success rate of 83 percent.
What sets this research apart, however, is not just the AI’s ability to perform tasks but also its ability to describe them in natural language to another AI. This “linguistic conversation” between AIs opens up new possibilities for collaboration and knowledge sharing in robotics.
The researchers created AI models with human-like learning and communication skills by simulating the areas of the human brain responsible for language perception, interpretation, and instructions-based actions. This breakthrough could have profound implications for various industries, from manufacturing to healthcare.
The researchers of this study envision building upon their work to create more complex AI networks to integrate into humanoid robots. One day, we may see robots capable of understanding and communicating with humans and other AI systems.