OpenAI debuts GPT-5, paving the way for an even smarter ChatGPT
Openai announced GPT-5 on Thursday, the fundamental model for the next generation of Chatppt and a “significant intelligence jump”, according to the company.
The GPT-5 will be published in two versions, a “pro” model which will only be accessible to paying subscribers, and the basic GPT-5, which will in fact be available for everyone, even those of the free Chatgpt plan. In addition, subscribers will have more use, said Openai.
Interestingly, Openai will not explicitly do which chatgpt model uses available for chatgpt users. An interesting feature of the new GPT-5 is a “real-time router” which will affect a model based on the request that the user requests. If you use the terms “think hard”, the chatgpt could see it as a signal that a more complex model should be used. An “effective” model can be used for most questions, with a deeper model of reflection available if necessary. Apparently, there is also a “mini” version that will launch if a user exceeds the use rates.
“We have made significant progress in reducing hallucinations, improving instructions to follow and minimizing sycophance, while improving GPT performance – 5 in three of the most common uses of chatgpt: writing, coding and health,” said Openai in a blog article.
A field of expertise is coding, where Chatgpt and GPT-5 can now create web and whole games with a single invite, said Openai, while exceeding many of its previous levels of expertise in health. Our sister site, Techcrunch, has much more on GPT-5.
