ChatGPT-4 users want GPT-5 to remove cap of 25 messages every 3 hours, cut reliance on plugins for math queries, & more

Artificial intelligence has made significant advancements in recent years, and one of the most notable developments is the advent of language models like GPT (Generative Pre-trained Transformer).

These models, such as ChatGPT-4, have revolutionized the way we interact with AI systems, enabling more natural and engaging conversations.

However, despite their impressive capabilities, there are still some limitations that users hope will be addressed in the next version, GPT-5.

One of the primary concerns raised by ChatGPT-4 users is the restriction on the number of messages that can be exchanged within a given time frame.

Currently, there is a cap of 25 messages every 3 hours, which can be quite limiting for users who engage in lengthy or complex conversations.

Yes. You have to start a new chat in GPT-3.5 if you send 25 messages in 3 hours. I think short and quick messages or very simple things sometimes don’t count. But I hit the limit all the time with complicated and long code debugging

Another area in which users hope to see improvement is in the handling of math queries.

While ChatGPT-4 is capable of performing mathematical calculations, it often relies on external plugins or tools to provide accurate results.

This can be inconvenient for users who expect the AI model to handle math queries without additional assistance. This will also make GPT-5 a more versatile and comprehensive tool for users across various fields.

Better at math without plugins
Source

yes, it’s a pain when coding. it completely blanks as soon as there Is any kind of math involved

In addition to these technical limitations, some users have expressed dissatisfaction with the multiple apologies generated by ChatGPT-4 whenever it fails to find the desired information or provide a solution.

While the AI model’s politeness is commendable, the excessive use of apologies can become repetitive and tedious for users.

Never ever say ” As an AI language model, I am unable/cannot/programmed to …” But instead just come up with the best compromise within its confines.
Source

This, and stop fucking apologizing for everything

Furthermore, one issue that has been raised by users is the tendency for ChatGPT-4 to forget previous parts of a conversation.

This limitation can disrupt the flow of a discussion and lead to repetitive or redundant interactions.

Not forgetting stuff mid conversation. The moment I can feed him books or a bunch of journal papers to work with in longterm memory is the day I rejoice.
Source

Interestingly, the problem of forgetting conversations is not unique to ChatGPT-4 alone. Character.AI, another AI platform, has also been reported to exhibit similar memory issues.

This indicates that the challenge of memory retention is a common concern among AI language models and underscores the need for further research and development in this area.

As the development of AI language models progresses, addressing what users want will undoubtedly be a priority for ChatGPT-4 devs in GPT-5 and future versions of it.

That being said, stay tuned to PiunikaWeb to know if and when there are any further developments regarding this topic.

PiunikaWeb started as purely an investigative tech journalism website with main focus on ‘breaking’ or ‘exclusive’ news. In no time, our stories got picked up by the likes of Forbes, Foxnews, Gizmodo, TechCrunch, Engadget, The Verge, Macrumors, and many others. Want to know more about us? Head here.