Machine Psychology: Unlocking Human-Level AI Potential

"Machine Psychology" Could Unlock Human-Level AI, Says Researcher from Linköping University . Artificial intelligence may soon achieve human-level intelligence, thanks to the emerging field of "Machine Psychology." This groundbreaking concept, developed by Robert Johansson in his Ph.D. research at Linköping University, combines psychological learning models with adaptive AI systems to approach the long-standing challenge of Artificial General Intelligence (AGI).

AGI, which could perform intellectual tasks like humans, has been a key goal of AI research since the 1950s. Johansson believes AGI could become a reality within the next five years, revolutionizing society by enabling AI to act as virtual researchers, psychologists, and more. Unlike current AI, AGI would embody a completely new type of intelligence.

Using the Non-Axiomatic Reasoning System (NARS), Johansson integrates psychological principles to mimic human-like learning and reasoning. This system operates with incomplete data and limited computational power, offering flexibility for real-world applications. Johansson describes AGI as transformative but emphasizes careful ethical and societal considerations as its development progresses.

Johansson’s approach has garnered attention from leading AI organizations like Google DeepMind. While AGI presents risks, such as societal misuse, Johansson envisions it as a tool for fostering understanding and countering global challenges.

Psychology will play a crucial role in unlocking human-level intelligence in AI, Johansson says, urging responsible development of this powerful technology.