OpenAI aims to develop AGI systems that can think like professional engineers

OpenAI is working on an artificial intelligence to help senior software engineers handle more complex programming tasks, according to three people familiar with the matter, a key step in the company's attempt to develop a general artificial intelligence (AGI) that outperforms humans at the most economically valuable jobs. The effort is part of Altman's goal of automating software engineering, according to a person who heard OpenAI CEO Sam Altman talk about the project. He hopes to release a powerful new product that can handle coding problems involving multiple steps and aims to replicate the work of experienced programmers. Such products are often referred to as "agents," and Altman recently said such systems could "join the workforce" as soon as this year. According to an employee, such agents could facilitate an internal tool that OpenAI has developed to help its artificial intelligence researchers work faster, including generating code for their experiments related to artificial intelligence models. OpenAI has been preparing to test early versions of advanced coding agents with a few specific prospects as it tries to make the product useful and popular, one of the people said.