OpenAI has officially announced the release of GPT-4.5, its newest and largest language model, which will be available as a research preview. According to the company, GPT-4.5 is its "largest and most knowledgeable model yet," boasting improved writing capabilities, enhanced world knowledge, and a "refined personality" compared to previous models.

However, OpenAI has tempered expectations by warning that GPT-4.5 is not a frontier model, meaning it doesn't introduce enough new capabilities to be considered a significant breakthrough. In a leaked document, OpenAI stated that GPT-4.5 "does not introduce 7 net-new frontier capabilities compared to previous reasoning releases, and its performance is below that of o1, o3-mini, and deep research on most preparedness evaluations."

Despite its limitations, GPT-4.5 has made significant strides in reducing hallucinations, a common issue in language models. According to OpenAI, GPT-4.5 hallucinates less than its predecessor, GPT-4, and slightly less than its o1 model. This improvement is likely due to the new supervision techniques and traditional methods used in its training, such as supervised fine-tuning and reinforcement learning from human feedback.

The release of GPT-4.5 is a significant step towards OpenAI's goal of combining its large language models to eventually create a more capable model that could be labeled as artificial general intelligence (AGI). The company's CEO, Sam Altman, has referred to the upcoming GPT-5 system as a "system that integrates a lot of our technology," which will include OpenAI's new o3 reasoning model.

It's worth noting that GPT-4.5 will be limited to ChatGPT Pro users, similar to how Operator was released. OpenAI is set to hold a stream to share more information on the release and availability of GPT-4.5. The company has also confirmed plans to launch GPT-5 by late May, which will further advance its AI capabilities.

The development of GPT-4.5 and its limitations highlight the ongoing challenges in creating more advanced AI models. While OpenAI's efforts are pushing the boundaries of language processing, the industry still has a long way to go in achieving true AGI. As the AI landscape continues to evolve, it will be interesting to see how OpenAI's latest release impacts the research community and the broader tech industry.