New Delhi, Nov 9 (NationPress) By the year 2026, artificial intelligence will be capable of making minor discoveries. However, OpenAI is optimistic that starting in 2028 and beyond, we will witness the emergence of systems capable of achieving much more significant breakthroughs.

The US-based AI organization notes that the majority of the global population still perceives AI primarily as chatbots and enhanced search functionalities.

“Currently, we possess systems that can surpass the intellect of the brightest humans in some of our most demanding cognitive challenges. Despite AI systems still exhibiting unpredictable behavior and encountering notable limitations, those that can tackle complex problems appear to be closer to 80% of the journey to an AI researcher’s goal rather than just 20%,” stated OpenAI.

The disparity between common AI usage and its actual capabilities is vast.

Furthermore, AI systems that can autonomously uncover new insights, or enhance human effectiveness, are poised to create a substantial impact on society.

OpenAI highlights that within a few short years, AI has progressed from executing tasks (particularly in software engineering) that a human could complete in mere seconds to those that typically require over an hour.

“We anticipate that soon, systems will handle tasks that would take a person days or even weeks; we are uncertain how to conceptualize systems capable of completing tasks that may take a person centuries. Concurrently, the cost of achieving a certain level of intelligence has significantly decreased, with a reasonable estimate being a 40x reduction annually over recent years,” they mentioned in a blog entry.

Very shortly, AI systems will assist individuals in understanding their health, boost advancements in areas such as materials science, drug development, and climate modeling, as well as broaden access to tailored education for students globally.

Highlighting these clear advantages helps cultivate a collective vision of a future where AI enhances life, not just efficiency.

“While the potential benefits are vast, we regard the dangers of superintelligent systems as potentially catastrophic, and believe that empirical research on safety and alignment can guide global decisions, such as whether the entire field should decelerate development to more thoroughly examine these systems as we approach those capable of recursive self-enhancement,” the organization remarked.

Clearly, deploying super-intelligent systems without the means to effectively align and control them is inadvisable, and achieving this necessitates further technical advancements,” they concluded.