OpenAI Japan CEO says incoming AI model GPT-Next will be 100 times better than GPT-4
At the recent KDDI Summit 2024, Tadao Nagasaki, CEO of OpenAI Japan, made a significant announcement regarding OpenAI's next AI model, dubbed GPT-Next. According to Nagasaki, GPT-Next will be 100 times more powerful than the current GPT-4, representing a massive leap in AI capabilities.
First hinted at during an event in Paris earlier this year, GPT-Next is poised to be the successor to GPT-4, promising enhanced multimodal abilities and smarter training models that could revolutionize AI functionality. Nagasaki mentioned that GPT-Next will utilize a smaller version of OpenAI's secretive Project Strawberry, a project that aims to create Artificial General Intelligence (AGI). This suggests that the improvements in GPT-Next will come from advancements in the model's architecture rather than just scaling up computing resources.
Nagasaki also shared impressive growth statistics, noting that ChatGPT had surpassed 200 million active users by the end of August, making it the fastest software to reach such a milestone. He emphasized the exponential rate at which AI is advancing, with OpenAI committed to accelerating AI's integration into everyday life.
The announcement builds on previous statements by OpenAI executives. Mira Murati, OpenAI's CTO, had earlier spoken about another model, GPT 4o, which was already a significant improvement over GPT-4 in terms of speed, cost, and capability. She also hinted that future models, likely including GPT-Next, could possess intelligence surpassing that of humans in certain tasks, marking a new era in human-AI interaction.
With GPT-Next on the horizon, the advancements promised by Nagasaki and Murati suggest a future where AI could become even more integral to various aspects of life, potentially outperforming human capabilities in specific areas.
