OpenAI CEO Sam Altman stated in a personal blog post that he believes OpenAI 'knows how to build (general artificial intelligence),' as it traditionally understands this and has begun to shift its goals towards superintelligence.
Altman wrote: 'We love our current products, but we are here for a brilliant future, where superintelligent tools can greatly accelerate scientific discovery and innovation, far beyond what we can do ourselves, thereby greatly increasing abundance and prosperity.'
AGI, or general artificial intelligence, is a vague term. But OpenAI has its own definition: 'Highly autonomous systems that outperform humans at the most economically valuable work.' OpenAI and its close partner and investor Microsoft also have a definition of AGI: an artificial intelligence system that can generate at least $100 billion in profit. According to the agreement between the two companies, when OpenAI achieves this goal, Microsoft will lose its rights to use the technology.
Altman did not explicitly state which definition he might be referring to, but he seems more likely to refer to the former. In the article, Altman believes that AI agents (AI models that can autonomously perform certain tasks) may 'join the workforce' and 'fundamentally change a company's output.'
This is possible, but today's AI technology also does have significant technical limitations, such as generating hallucinations and making errors. Altman seems to believe that this can be overcome, and at a relatively fast pace, but the timeline is unpredictable.