According to the business, their technology is getting close to reaching the second stage of the five stages of artificial general intelligence.
New Delhi, 12th July 2024
In an effort to shed light on its views on safety and the future of artificial intelligence, OpenAI has developed a five-level structure to track its progress in developing software that can perform tasks that humans cannot.
During an all-hands meeting on Tuesday, OpenAI, which is widely recognized as a leader in the development of cutting-edge AI systems, presented a new classification system to its staff, a spokeswoman said.
The levels, which OpenAI plans to provide to investors and outside stakeholders, range from the present conversational AI capabilities (Level 1) to the potential future AI that may carry out organizational duties (Level 5). The purpose of this project is to make the company’s AI development objectives and their effects on the larger community more transparent.
Also Read : Itarhi Farmers Producer Company Gives Bihar a Special Recognition in the Field of Agricultural Products
As to the statement provided by a representative, OpenAI executives told staff members that the organization is now categorized at the first level but plans to go up to the second level, which is called “Reasoners.” This phase describes systems that, in the absence of external aids, are able to perform fundamental problem-solving activities on par with a human physician.
A person with knowledge of the discussions who asked to remain anonymous said that business executives presented a research project showcasing their GPT-4 AI model during the same conference. OpenAI claims this model displays new skills that approach human-like reasoning. An industry norm, internal exploration of new capabilities is what OpenAI consistently does, an industry spokesperson said when questioned about the presentation.Although there are currently no artificial general intelligence (AGI) systems, OpenAI has long claimed that one of its goals is to create one. AGI systems are computers that can do better than humans on the majority of activities. The CEO of OpenAI, Sam Altman, has voiced optimism that artificial intelligence (AGI) will be achieved in this decade. There has been continuous discussion among AI researchers over the requirements for reaching AGI.
Google DeepMind researchers unveiled a framework with five stages of AI advancement—including “expert” and “superhuman”—in a study that was released in November 2023. This grading scheme is similar to the method used by the auto industry to assess the degree of automation in self-driving vehicles.
Also Read : Over 100 Non-Certified Toys Seized in Ghaziabad BIS Raid: Know Here
