Over the past weekend, the artificial intelligence (AI) community has been abuzz with discussions on social media about a possible upgrade to the popular AI chatbot ChatGPT. This speculation gained traction following the circulation of an image that purportedly revealed a new version of ChatGPT, labeled as GPT-4.5. This version is rumored to include enhanced multimodal capabilities and a revised pricing structure, as of December 14.
Despite these rumors, OpenAI CEO Sam Altman has publicly refuted claims about the immediate release of this new update for the Microsoft-supported AI chatbot. In contrast, industry observers have continued to report further supposed ‘leaks’ of the GPT-4.5 model. One such observer, Rowan Cheung, noted an upsurge in the performance of the existing model over the last two days, fuelling further speculation among followers.
Queries to ChatGPT about its current model and API have yielded unusual responses. OpenAI’s Will Depue referred to these as a “consistent but strange misconception.” In one instance, ChatGPT identified its running model as ‘GPT 4.5-turbo,’ further intensifying the intrigue around its capabilities.
AI Sector: Competitive Advances and Ethical Developments
In recent days, several reports have claimed that ChatGPT exhibited marked improvements, performing better than it ever has. Additionally, the ChatGPT X social media account has been posting enigmatic messages, adding to the speculation.
Despite Altman’s denials, Cheung suggests that some form of upgrade to ChatGPT might be in progress, even if a full-scale release of a new version isn’t imminent.
In broader AI news, OpenAI and ChatGPT are racing to match strides with rivals like Google and Gemini. Google unveiled its advanced multimodal chatbot, Gemini, on December 6, touted as a “ChatGPT killer.” However, the credibility of Gemini faced a setback after revelations that its demo video was fabricated.
Meanwhile, the AI sector is witnessing other significant developments. Sacramento State recently established the National Institute for Artificial Intelligence in Education, focusing on the ethical application of AI in educational contexts. French AI company Mistral launched its new large language model, Mixtral 8x7B, claiming superior performance over GPT-3.5 and Meta’s Llama 2 family.
In China, AI giant SenseTime’s shares have plummeted to their lowest point following the demise of founder Tang Xiao’ou on December 15, at 55 years old. His death has sent ripples through the industry, highlighting the impact of individual visionaries on the rapidly evolving field of artificial intelligence.