Sam Altman Wants To Make AI Like A ‘Super-Competent Colleague That Knows Everything’ About Your Life

In Sam Altman’s vision, artificial intelligence (AI) transcends its current capabilities to become a super-competent colleague, as he expressed in an interview with the MIT Technology Review.

Altman’s ideal AI counterpart is an intelligent being with a profound understanding of the user’s life but without the usual “extension” feeling. This AI should be self-starting, able to work quickly and independently, but also intelligent enough to ask questions when it needs them answered. Altman highlights the need for AI to go beyond being a simple chatbot and instead try to impact jobs performed in the real world significantly.

However, Altman’s vision for AI goes beyond what OpenAI has to provide. Even while ChatGPT and other current tools are used for job completion and workflow acceleration, Altman has called them “incredibly dumb” compared to his envisioned AI colleague. Nevertheless, Altman believes that when this cutting-edge AI model hits the market, productivity will go up substantially.

Although Altman does not provide a timeframe for deploying this tool or the necessary advances in AI, OpenAI appears to be making headway. GPT-5, the upcoming language model, is a promising first step towards Altman’s goal.

According to reports, GPT-5 shows appreciable improvements over current models and could lead to a service that lets customers assign jobs to AI agents independently.