Folks, OpenAI has developed a five-level scale to assess the development of AI capable of surpassing humans Let's see what these levels are and when we should be scared?
The first level, Chatbots— means the ability of a neural network to interact with people "in a spoken language." According to the company, its AI is now on the threshold between the first and second levels called Reasoners. To fully achieve this goal, the neural network must perform basic tasks at the level of a person with a doctoral degree (PhD, folks!) who does not have access to any tools.
The third level, Agents, assumes that the AI can perform actions on behalf of the user for several days. At the fourth level, Innovators, the neural network is capable of inventing innovations The fifth level, Organizations, is characterized by the ability to perform the work of an entire organization.
When do you think we'll reach level 5? 🤖
