In March of this year, OpenAI unveiled GPT-4 as the successor to ChatGPT. Fresh reports now indicate that GPT-5 could be in development and on its way
In November 2022, OpenAI’s ChatGPT opened up new possibilities in the world of artificial intelligence (AI). While AI itself was not a new concept, ChatGPT’s immense popularity sparked a surge of interest in generative AI. People marvelled at how effortlessly this AI chatbot could assist them with various tasks, from coding and poetry composition to tackling complex questions and excelling in exams. However, concerns about potentially negative aspects also emerged.
Merely six months after the debut of ChatGPT, powered by the formidable GPT-3.5 language model, OpenAI introduced GPT-4—an even stronger, smarter, and more potent iteration of its predecessor. Now, recent reports indicate that OpenAI might be on the brink of launching GPT-5, fueling further excitement and anticipation in the AI community.
Is ChatGPT-5 coming soon?
As reported by Windows Latest, OpenAI has filed a trademark application that strongly implies the existence of GPT-5. The application, discovered by the publication and dated back to July 18, indicates that OpenAI is likely preparing to introduce another substantial language model in the near future.
According to the latest report from Windows Latest, OpenAI has submitted a new trademark application to the USPTO for ‘GPT-5’, describing it as ‘downloadable computer software for using a language model.’ Interestingly, similar terms were used in the trademark applications for GPT-3.5 and GPT-4.
The mention of ‘GPT-5’ in the application has sparked speculation that OpenAI may be working on an even more potent version of ChatGPT. However, no additional information has been disclosed, and it’s worth noting that the launch of GPT-5 might not necessarily happen this year.
An official confirmation from the OPenAI team is still pending
Regarding ChatGPT 4
OpenAI’s blog post introducing GPT-4 stated, “We’ve developed GPT-4, the most recent milestone in OpenAI’s endeavour to scale up deep learning. GPT-4 is a substantial multimodal model capable of processing both image and text inputs and producing text outputs.. Although it may be less capable than humans in certain real-world situations, it demonstrates human-level performance on various professional and academic benchmarks.”
During its launch, OpenAI provided a list of exams that GPT-4 successfully passed, along with the corresponding scores. The language model achieved impressive results, such as an 88th percentile on the LSAT and an 89th percentile on SAT Math. Additionally, it scored in the 80th percentile on the GRE Quantitative exam, and remarkably, attained the 99th percentile on the GRE verbal exam and the 54th percentile on the GRE writing exam.