Sam Altman says AGI has become a pointless term — experts agree
Openai CEO Sam Altman speaks on June 2, 2025 during the snowflake summit in San Francisco.
Justin Sullivan | Getty Images News | Getty Images
Openai CEO Sam Altman said that artificial general intelligence or “AGI” has lost interest as a term because it makes it difficult to define the concept of rapid progress in the field.
AGI refers to the concept of an artificial intelligence that can perform any intellectual task that a person can do. Openai has been trying to investigate and improve AGI for years, which has been safe and benefits all humanity.
When Altman asked CNBC’s “Squawk Box” last week, when the company’s latest GPT-5 model was closer to reaching the world AGI, “I think this is not a super useful term.” He said. Artificial intelligence entrepreneur said that AGI had previously thought that it could be developed “reasonablely in the near future”.
The problem with the AGI is that there are multiple definitions used by Altman, different companies and individuals. According to Altman, a definition is an AI that can “do a significant amount of work in the world – but have these problems, because the nature of the work changes constantly.
Altman said, “All this is not really important, and this ongoing folder of the model capability that we will trust only for more things.” He said.
Altman, “AGI” and how people use the term is not alone in skepticism.
Difficult to define
Futurum Group Vice President and AI Training Presidency Nick Patience, CNBC’ye AGI’ye “a fantastic north star for inspiration”, although it is not a useful term in general, he said.
“He directs the fund and captures the public imagination, but the definition of sci -fi creates a hype fog that usually hides the real, concrete progress we make in a more private AI.” He said.
Openai and other initiatives have gathered billions of dollars, and eventually reached high values with the promise that they would reach a type of AI enough to be considered as “net”. Openai was valued by the latest investors worth $ 300 billion and a Secondary shares sales in a valuation of $ 500 billion.
Last week, the company released the latest GPT-5, the latest major language model for all CATGPT users. Openai said that the new system is smarter, faster and “much more useful” – especially when it comes to writing, coding and helping in health queries.
However, the launch led to criticism that the long -awaited model was an overwhelming upgrade and that it only provided small improvements on its predecessor.
“All accounts are increasing, not revolutionary,” said Wendy Hall, Professor of Computer Science at the University of Southampton, said.
Hall, AI firms, “Globally agreed metrics should have to declare how they have to declare” added. “The wild west for snake oil sellers.”
A distraction of attention?
Altman accepted Openai’s new model Missing the sign Since the system has not yet learned on its own, from its own definition of personal network.
Openai still maintains artificial general intelligence as its ultimate goal, altman said it was better to talk about the levels of progress towards this overall intelligence rather than asking if something is AGI.
“Now we are trying to use these different levels …` “ “ or not? ‘I think this was very rude as I approached, “Openai CEO he said during a conversation At the Finreglab AI Symposium in November 2024.
Altman is still waiting for AI to make some important breakthroughs in certain areas such as new mathematical theorems and scientific discoveries in the next two years.
“The very exciting real world is happening, I feel that AGI is a little distraction, it has been supported by those who should continue to create a surprising amount of financing.”
“Talking about certain abilities is more useful than this uncertain ‘general’ concept of intelligence.”
