Summary
- OpenAI launched GPT-5.1 on ChatGPT, promising a smarter and more pleasant assistant, with eight personality options, such as “Professional” and “Different”.
- GPT-5.1 has two versions: “Instant”, which is more welcoming and efficient, and “Thinking”, which understands better and is faster in simple tasks.
- The model will initially be accessible to subscribers of the Business, Pro, Plus and Go plans, while free users will have access after the first distribution stage.
OpenAI updated ChatGPT this Wednesday (12/11), which now has the GPT-5.1 model. The company promises that the robot is smarter and more pleasant, balancing friendliness and efficiency. In addition, the assistant will have eight personality options, ranging from “Professional” to “Different”.
The new artificial intelligence model comes in two versions. The GPT-5.1 Instant is, in the company’s words, “more welcoming, smarter and better at following instructions” than its predecessor. GPT-5.1 Thinking “understands better and is faster in simple tasks, and more persistent in complex work”.
The default option leaves it up to ChatGPT to decide which model best suits the prompt, but the user can also choose which one they want to use.
GPT-5.1 will begin to be released in the next few days, reaching first to payers of the Business, Pro, Plus and the recently launched Go plans. Those who use it without paying or logging in will only have access to the model when this first stage of distribution ends.
Models based on GPT-5, launched in August, will be available to subscribers for another three months, in the ChatGPT menu. After that, they will disappear from the options.
ChatGPT gains new response options
ChatGPT already had some personality options, which change how the robot generates responses. The list has changed and now has more alternatives: Standard, Professional, Friendly, Frank, Different, Efficient, Nerd and Cynical.
OpenAI will also carry out, with a limited number of users, tests involving more specific controls over the format of responses, with adjustments to conciseness, friendliness, readability and even the use of emojis. ChatGPT itself may suggest updates to this configuration during conversations.
As much as they may seem less relevant, these changes in tone and personality tend to be more important than previously imagined for ChatGPT users. During the migration from GPT-4o to GPT-5, there were complaints about short and “impersonal” responses.
Before that, OpenAI made a mistake in the opposite direction: an update to GPT-4o transformed the chatbot into a “suck-up”, with an excessively positive tone, avoiding confrontations and using expressions such as “excellent question” and “I loved the depth”.
It is worth saying that this personification of ChatGPT (and technology as a whole) is the target of criticism. Mustafa Suleyman, from Microsoft, has already stated that tools that appear conscious are dangerous, for example. Cases of psychosis involving AI and lawsuits against OpenAI for allegedly encouraging suicide also had negative repercussions for the company.
With information from OpenAI and 9to5Mac
Source: https://tecnoblog.net/noticias/openai-libera-gpt-5-1-no-chatgpt-e-promete-assistente-mais-caloroso/
