OpenAI's "GPT-6" model, called "ChatGPT Astra," was released last week and achieved striking results in benchmark tests that seemed to distinguish it from rivals like Anthropic's Claude and Google's Gemini.
OpenAI said the model is "at the highest available level in computer use, web browsing, software engineering, cybersecurity, science, and professional work" and is capable of performing multiple tasks simultaneously and autonomously.
In a demonstration, the new AI model handled user requests while simultaneously performing complex tasks such as building a video game or preparing a legal document.
Huang, after unveiling the model, wrote on X: "General AI has arrived. Congratulations to the OpenAI team."
The precise definition of "general artificial intelligence" is disputed among figures in the industry, although according to OpenAI's charter, the concept includes "highly autonomous systems that outperform humans on most economically valuable work."
Greg Brockman, president of OpenAI, described the new ChatGPT release as "a generational leap in capability" in a conference call with reporters and said, "Welcome to the era of general AI."
He added: "If you jump forward a few years and then look back and ask when general AI was really built, I think the answer will be around this time, and I think it might be this model."
His claims came just days after OpenAI CEO Sam Altman said that general AI is "at best... a term with a very vague definition."
On the "Sources" podcast he said: "I wanted to say that this definition of general AI is kind of a meaningless marketing term."
In independent benchmark tests aimed at measuring how close a model is to general AI, ChatGPT Astra posted newly high scores that were very close to human scores.
Greg Komrat of the ARC Prize foundation, which develops these tests, described it as "a meaningful change in the performance of leading models."
Nonetheless, some industry figures remain skeptical about calling any current AI models human-equivalent.
Marco Tasic, a technology consultant, in response to Huang's remarks, rejected his claims and described them as a marketing stunt ahead of a potential initial public offering (IPO).
He said: "No, this is not general AI. This is just pre-IPO public relations. Artificially inflating the price and then dumping it, nothing more."