“What you really want,” the OpenAI CEO told the MIT Technology Review, is a “super knowledgeable colleague who knows absolutely everything about my life, every email, every conversation I’ve had, but which doesn’t feel like an extension.”
And they are self-sufficient and don’t need constant direction. They will instantly tackle some tasks, probably the simplest ones, Altman said. They will do a first try on more complex tasks and come back to the user if they have questions.
Ultimately, Altman wants AI to function as more than just a chatbot. It should help people accomplish things in the real world, he said.
This would be a considerable step forward from what OpenAI currently offers.
Altman reportedly called ChatGPT “incredibly stupid,” even though workers are already using it to speed up their workflows, develop code, write emails, and more. So it’s unclear how much more productive we’ll become once Altman’s fellow magic model hits the market.
Altman did not say when this tool will be available and how advanced the AI needs to be to support it. The company’s other offerings, like the Sora Video Generator and the DALL-E Image Generator, still require considerable guidance to accomplish their tasks. They are also not designed to perceive information from the environment and use it to achieve specific goals.
But OpenAI’s next language model, GPT-5, could be a step in that direction.
A source who has seen it previously told BI that it is “materially better” than existing models. The source also said that OpenAI is developing a service that allows users to call an AI agent to perform tasks autonomously.
Sources said GPT-5 could be released mid-year. Altman, however, doesn’t say much.
“Yes,” he simply told reporters this week at an event in Cambridge, Massachusetts, where he was asked when OpenAI would release GPT-5.
businessinsider
