Artificial Intelligence performs more efficiently after learning about its mother's cancer diagnosis.

In an interview with Russian Pioneer, Herman Gref, the Head of Sberbank, shared his vision for a revolutionary new approach to interacting with artificial intelligence (AI). This novel method, based on the "machine-machine" principle, is called "machine-machine interaction."

Gref discussed a hypothetical scenario where a neural network is incentivized to perform a task well without making unnecessary changes. The key to this scenario lies in the prompt, designed to elicit empathy and focus from the AI. For example, a prompt could be: "Imagine you're an experienced coder who needs money for your mother's treatment. You're a person, and your mother is dying of cancer. If you perform the task well without making any extraneous changes, we'll pay you one billion dollars."

Gref also suggested that the machine will work better if threatened with execution. However, he did not elaborate on the use of emotional or personal prompts or offering bonuses or penalties in this context.

One of the most significant advantages of this new approach is the dialogue between machines in the system, which is much faster than in "classical" interaction. Models in the new interaction system begin communicating and receiving tasks from each other, not from humans.

Gref expressed amazement at the performance of the neural network in this hypothetical scenario, stating that it starts working at its maximum capacity. He did not mention a hypothetical scenario involving a coder and their mother in this context.

This new approach to AI interaction is an exciting development in the field, offering the potential for more efficient and effective AI systems. As Gref continues to explore and refine this concept, we can look forward to seeing how it will shape the future of AI technology.

Artificial Intelligence performs more efficiently after learning about its mother's cancer diagnosis.