One of the leaders in the race to achieve general artificial intelligence, OpenAI developed a five-tier system to track progress toward this goal. The company recently released this rating to its employees and plans to share it with investors and others stakeholders in the future.
Currently, OpenAI considers its AI systems to be at the first level, which refers to the ability for conversational interaction. This level, called “Chatbots“, describes the stage at which AI is able to maintain complex and natural dialogues with humans, answering questions and performing basic text-based tasks.
The company says it is approaching the second level, called “Reasoners,” which involves solving complex problems, comparable to a human with a PhD but without access to tools. At this stage, the AI should be able to handle tasks that require advanced reasoning and critical thinking, solving problems based on a deep understanding of the data and context presented.
The third level, called “Agents,” describes systems capable of performing different actions on behalf of a user over several days. These agents would have the ability to perform continuous and complex tasks, such as scheduling appointments, managing projects, and even making decisions on behalf of users, operating in a nearly autonomous manner.
The fourth level, called “Innovators,” refers to AI that can help develop inventions. At this stage, AI systems would not only perform tasks but also actively contribute to the creation of new technologies and innovative solutions, working alongside human scientists and engineers to accelerate progress in a variety of areas.
The fifth and final level, called “Organizations,” will be reached when AI can perform the work of an entire organization. This would include managing complex business processes, from day-to-day operations to high-level strategic decisions, allowing a company to function efficiently and effectively with minimal human intervention.
The ranking was developed by OpenAI executives and may be adjusted based on feedback employees, investors and the board. In May, the company disbanded its “Super Alignment” team, moving members to other internal research projects in a move that aims to streamline efforts and resources to achieve AGI.
Towards the future
Recently, OpenAI CEO Sam Altman and CTO Mira Murati expressed confidence that AGI will be achieved within the next 10 years. Altman stated that the creation of AGI will be one of the most powerful tools ever developed by humanity, enabling significant advances in several areas of knowledge and technology.
In March this year, Nvidia CEO Jensen Huang also expressed his belief that AGI could be achieved within five years, if it is defined as the ability to pass any proposed tests. Huang highlighted that the rapid evolution of AI technology and the increase in computing power are crucial factors in achieving this ambitious goal.
Elon Musk has suggested that AI could surpass the most intelligent humans by 2025 or 2026, mentioning that the total amount of “sentient computing” would surpass that of all humans within five years. Musk emphasized the importance of developing AI responsibly, ensuring that its impact is beneficial to society as a whole.
* With information from QZ and Bloomberg
Follow Adnews on Instagram e LinkedIn.
#ADDSomethingNew #ADNewsOficial #ADGroupBR
