OpenAI carried out a survey among the users of the find out how to improve your ChatGPT. The result? A mix of striking the technical requirements and the improvement of the practice will make a chatbot in a more powerful and more comfortable for you to use in your daily life.
common themes:
AGI
agents
much better 4o you
much better memory
longer side context
“grown-up”mode
deep research feature
better sora
more personalization(interestingly, many great updates we have coming that were mentioned not at all or very little!) https://t.co/lMZmlZif66
— Sam Altman (@sama) December 30, 2024
What's new from the OpenAI to 2025: and, lo, thus requiring the user
At the top of the list is the AGI (the intelligence of a total artificial). Some can only dream of for the one that can reason and learn as a person, or even better. This vision inspires both excitement and even fear, especially with regard to the issue of the ethics and safety. However, OpenAI does not seem that there is in front of this challenge titanic, if the path to AGI is still full of pitfalls.
After the AGI, the users are interested in, to the agents of the IT, e.m.th. bots are automated to be able to take the initiative. For example, the ChatGPT you can manage your calendar, reserve tickets and train, and organize documents, all without the intervention manual. This feature, referred to already in the OpenAI, promises a significant step forward in the quality, in terms of helping the users in their daily activities.
Select “and " mode of education”
The lovers, after scientific research, and the profession expect for an item that the search of the depth along the line of the Search of the Depth of Gemini, allowing ChatGPT to be digging a lot deeper than just an answer to a single. At the same time, the speed is another important challenge: to integrate the preferences of each user in order to provide suggestions as to the most relevant, or even a style of response to be appropriate for their field of activity.
Based on the request's “the mode of education”. The idea behind it? Facilitate this activity, and let the ChatGPT to explore topics in a more vulnerable, even to the issue. In other words, in a way, with little or no censorship. However, it is difficult to imagine that the OpenAI of çaktivizojë to fully protect the current, taking into account the reputation and security of the platform are at risk.
Any longer, and the memory of the best
A critique of a repeat of it has to do with the memory limit of the ChatGPT. When you start a chat in the chatbot and sometimes it tries to retain the money. The user would like larger capacity of memory, and a story to be really useful, so as to avoid repetition and unnecessary. Therefore, the OpenAI is explore that, as well as to allow for the model to retain a lot of information on kontekstual, which can improve significantly the flow of the discussion and the relevance of the answers.
An improved version of Sora
Sora is definitely one of the generators are the best videos with IT, but in the process of its long-term development, it has allowed people to pick up the pace. Although it may seem unthankful days of the request for a “Sora greater good,” so soon after the release of its rivals, as the Point was to steal the show.
Discussion about this post