OpenAI could launch new 'GPT-4.5 Turbo' AI model in June
What's the story
Several social media platforms were recently flooded with discussions about an all-new AI model from OpenAI, the creators of ChatGPT, called GPT-4.5 Turbo. The buzz originated from a blog post that was seemingly published by mistake and later removed by the company. However, search engines such as Bing and DuckDuckGo had already indexed the post, leading to its discovery by vigilant users on Reddit and X.
About the update
GPT-4.5 Turbo is touted as a faster, more accurate model
The leaked blog post said that GPT-4.5 Turbo will outperform its predecessor, GPT-4 Turbo, in terms of speed and accuracy. The description also mentioned a "knowledge cutoff" of June 2024, leading to speculation about a potential release date for the new model. Additionally, the post revealed an increased token context window of 256k tokens for the upcoming model, doubling the capacity of its predecessor.
Speculation and celebration
OpenAI has yet to comment on the alleged leak
Despite the widespread speculation, OpenAI has yet to comment on the alleged leak or confirm the existence of GPT-4.5 Turbo. The timing of this incident is noteworthy as it coincides with the first anniversary of GPT-4's launch on March 14. The company's silence has left many questions unanswered, adding to the intrigue surrounding this potential new AI model.
Other upgrades
Microsoft Copilot upgraded to GPT-4 Turbo
In related news, Microsoft's AI chatbot, Copilot, has been upgraded to the GPT-4 Turbo model. The announcement was made by Mikhail Parakhin, CEO of Advertising and Web Services at Microsoft. Both free and Pro users can now use the latest model, but only Pro users have the option to switch back to the older GPT-4 model. This upgrade is expected to enhance the user experience with more updated responses due to an increased knowledge base.