They update the paid ChatGPT so that it takes fewer detours and gets to the point

0
12
they update the paid chatgpt so that it takes fewer detours and gets to the point
they update the paid chatgpt so that it takes fewer detours and gets to the point

OpenAI has introduced changes to its more advanced ChatGPT model to make it less pompous and more precise in its responses. Accustomed to giving extensive answers with context and warnings about their own limitation, this could be a step towards an AI language closer to human.

Those who can afford the premium version of the well-known chatbot will be able to save time when dealing with ChatGPT. OpenAI has updated its GPT-4 Turbo model to be more direct, “ less verbose , and use more conversational language.”

Everyone who has poked around with ChatGPT to date will have noticed: the AI ​​model is usually very correct and provides plenty of warnings, clarifications or contextualizations every time we ask it something. There are those who will appreciate that AI is somewhat humble in this sense and avoids giving categorical answers. However, it can be annoying if a one-sentence answer is enough.

It seems that its developers are aware that all this verbiage is excessive in some cases, and they have taken this into account when creating the latest version of this product, called “gpt-4-turbo-2024-04-09”. Unfortunately, at the moment only users with a ChatGPT Plus, Team or Enterprise account will be able to experience this new face of the chatbot that is less prone to verbosity.

This was announced by OpenAI in X this Friday, sharing a screenshot of the difference in length between a response following the old operation and another from the new version:

 

«Our new GPT-4 Turbo is now available for paying ChatGPT users. “We have improved skills in writing, mathematics, logical reasoning and coding,” the company said. This GPT-4 has also expanded its knowledge base, which is now limited from April 2023 to December 2023. In the post, OpenAI also attaches a graph that illustrates these improvements in several indicators.

Last Tuesday, OpenAI updated its API to introduce new models such as the GPT-4 Turbo with Vision.

The artificial intelligence company continues its path, but not without having gone through some bumps, after learning that Microsoft tried to sell its text-to-image generator DALL-E to the US military, or that two researchers have been fired for allegedly leaking information.

More natural language for ChatGPT

On the other hand, more colloquial language could make ChatGPT sound much more realistic, so that it seems like you are actually talking to another person. Those who use the language model not so much to obtain answers, but to maintain conversations may appreciate this change that more humanizes the chatbot.

Although it is true that there are tastes for everything. In responses to OpenAI’s announcement on X, some users express discontent with the update. «Something changed, that’s clear. After work he wanted to talk about history and instead of giving me interesting dialogue, he kept giving me phrases in response :(«, points out user @Dima_FilFabric. «Yes. I don’t like it. It’s like going back to the old days of GPT-3, that is, short and empty answers because it is a small mode,” says @FreudianHIPS.

More information about the code behind this latest version is available on OpenAI’s GitHub .

Previous articleThis widget comes installed on your mobile: it is one of the most useful you can find
Next articleMovistar seeks to take advantage of the merger by negotiating with Masorange the purchase of its 5G frequencies
Abraham
Expert tech and gaming writer, blending computer science expertise