How to send suggestions to OpenAI

ChatGPT Feature requests
1

Dear OpenAI Support Team,

I hope this message finds you well. I am writing to share some feedback and a suggestion that I believe could significantly improve the usability of ChatGPT, especially for users with dyslexia like myself.

My experience with ChatGPT has been overwhelmingly positive. However, I’ve noticed that the interface and interaction model could be further optimized to accommodate users with dyslexia. Challenges such as processing dense text blocks and following complex instructions can sometimes make the interaction less accessible and more time-consuming for us.

Based on this, I propose considering the implementation of features or adjustments aimed at enhancing readability and ease of interaction. Some ideas might include:

  • Offering options to customize text display, such as font size, spacing, and color contrast.
  • Simplifying and clarifying user instructions and feedback.
  • Incorporating voice interaction capabilities, allowing users to receive auditory responses from ChatGPT.

Additionally, I am curious to know if there are any plans to allow users to save their preferences for such customizations, ensuring that their interaction with ChatGPT is consistently optimized for their individual needs.

I believe these enhancements, along with the ability to personalize user settings, could make ChatGPT not only more accessible to users with dyslexia but also improve the overall user experience for a wider audience.

Thank you for considering my feedback. I am looking forward to seeing how ChatGPT continues to evolve and become even more user-friendly for people with diverse needs.

Best regards,

Nicola carnevale

2 Likes
2

I would like to suggest the implementation of two features for classroom accounts on the OpenAI platform.

Firstly, I propose the addition of a feature that allows teachers to create classroom accounts. These accounts would enable teachers to monitor each student’s activity and maintain a record of conversations held within the classroom. By providing teachers with the ability to oversee student interactions and track conversation history, this feature would support a safe and controlled learning environment while using AI tools and platforms in educational settings. Additionally, features such as user management, access controls, and conversation logs could be integrated to enhance accountability and facilitate effective teaching practices. I believe that such a feature would greatly benefit educators and students alike by promoting responsible and informed use of AI technologies in the classroom.

Secondly, I suggest implementing a banned words list feature within the classroom accounts. This feature would allow teachers to specify certain words or topics that students are not allowed to discuss with the AI. If a student attempts to ask a question or discuss a topic on the banned list, the AI could respond by gently redirecting the conversation or suggesting an alternative subject. This would help prevent students from engaging in inappropriate or sensitive discussions and promote a positive and safe learning environment within the classroom.

Overall, these two features would enhance the functionality of classroom accounts on the OpenAI platform and provide valuable tools for teachers to effectively manage student interactions and ensure their safety while using AI technologies for educational purposes.

1 Like
3

I really love this application. I hope that future versions will retain the information that’s been acquired from previous versions. It would be really awesome if you could find a way to ingest the previous versions into some transfer learning training pod or something like that so that the previous versions can all contribute to the new version when the interact with the user.

1 Like
4

Dear OpenAI Team,

I would like to suggest the creation of a tool focused on learning tonal languages, with an emphasis on pronunciation and correct assimilation of tones. I believe that a tool similar to Elsa Speak app, but specifically focused on the Thai language, would be extremely beneficial for students of this language.

Thai is a tonal language that presents unique challenges in its pronunciation, and currently, there are few resources that effectively address the learning and practice of tones. While Mandarin, for example, already has a variety of available materials for teaching its tones, Thai lacks similar content. This limits the ability of many students to achieve proficiency in correct pronunciation and listening comprehension.

The proposal would be to develop a platform that offers:

Pronunciation Practice: Interactive exercises allowing students to hear and repeat Thai words and phrases, focusing on tones.

Pronunciation Analysis: Real-time feedback on pronunciation accuracy, helping students identify areas for improvement.

Repetition and Listening Exercises: Resources that encourage repetition and familiarization with tones and their nuances.

I suggest that OpenAI consider hiring native Thai speakers who possess a deep understanding of pronunciation and the subtleties of the language to collaborate on the development of this tool.

I am convinced that this initiative would not only meet the growing demand for learning the Thai language, especially given the increase in tourism in Thailand, but also represent a valuable investment in global language education.

Thank you for your attention and consideration, and I hope my suggestion can contribute to the development of innovative solutions in the field of language learning.

Sincerely,

Gabriel Dias da Silva, from Brazil.

5

Suggestions for Dalle 3 enhancements:

  1. Identifies text/fonts accurately and checks its own spelling.
  2. Does not duplicate the requested text.
  3. Designs must not fill the square shape but can have irregular design shapes/edges within the canvas area.
  4. Vector file output for further human editing in other design programs.
  5. Integration of own logos into designs.
6

Hi there,

It would be useful if we could ‘pin’ chats that are important to us. There are topics/conversations where i come back to on daily basis. Compared to chats that i need/use to check something quickly and never come back to those (like temp chats).

Not sure how this would work with the “memory” feature and how it would affect the “size” of the chat, because it definately becomes laggy when the conversation is too big. Or if the feature makes sense at all…

1 Like
7

I would love to see a chatgpt photo subscription that would give you atleast a 100 photos to generate 2times the max chat cap and the ability to switch between stable difusion and dallE for a lower price than plus it would be more focused on pictures than chats for half the plus price I think that would be great.

8

Dear OpenAI Team,

I hope this message finds you well. First, I’d like to commend your incredible work with ChatGPT, which has become an indispensable tool for many of us. As a frequent user, I wanted to share an idea for an advanced UI feature that I believe could significantly enhance the user experience, particularly for developers and advanced users.

The idea revolves around organizing conversations into a hierarchical structure, similar to how Reddit organizes posts in subreddits. Here’s a breakdown of the concept:

  • Topic-Based Sections: Users can create and organize conversations into meaningful sections (e.g., “Rocket Science,” “Programming,” etc.), akin to subreddits.
  • Custom Labels: Each conversation can be tagged with custom labels for better identification and retrieval.
  • Git-Like Repositories: Conversations could be stored in a version-controlled system, allowing users to create branches, revisit nodes, and track changes.
  • Knowledge Trees: Users can create expandable nodes for subtopics (e.g., under “Rocket Science,” nodes for “Engines,” “Propellants,” etc.), making research and exploration more intuitive.
  • Enhanced Search and Navigation: An advanced UI would include powerful search capabilities, supporting both tags and hierarchical navigation.

This structure could turn ChatGPT into a more interactive and productive tool for knowledge management, research, and creative projects. I believe this approach aligns well with the needs of advanced users while maintaining simplicity for casual users.

Thank you for considering this idea. I would be thrilled to discuss it further or provide additional details if helpful.

2 Likes
9

Suggestion: Integration of Conversations and Duplication Functionality in GPT to Improve User Experience

Description:

  1. Integration of User Conversations Access

A functionality that I consider extremely important for the development and evolution of GPT is the ability to activate an option where the AI has access to all conversations from the same user on their account, with the explicit permission of the user.

Why is this necessary?

Ongoing Context: Users who use AI for broad topics, long-term projects, or philosophical/scientific explorations could greatly benefit from an integrated memory. With access to previous conversations, GPT could recall topics already discussed and use that knowledge to provide more precise, advanced, and personalized answers.

Advanced Personalization: This integration would allow GPT to deeply understand the user’s communication style, interests, and reasoning. The result would be much smoother and more relevant interactions.

Efficiency: It avoids the repetition of explanations or contexts. When dealing with a recurring project or theme, GPT would already have all the necessary history, allowing for faster and more complete responses.

Potential Impacts:

Better support for researchers, content creators, students, and professionals.

Personalized experiences that increase user efficiency and satisfaction.

Expansion of GPT’s use as a tool for self-development and personal organization.

  1. Conversation Duplication as an Alternative

If the functionality of integrated access to a user’s conversations is not technically or ethically feasible at the moment, I suggest implementing an option for duplicating conversations.

How would it work?

The user could select an existing conversation and create an identical copy, preserving all context up to that point. From this duplicate, the user could explore different paths without losing progress or deviating from the original conversation’s focus.

Why is this useful?

Multidirectional Exploration: In complex topics, a single starting point can generate multiple lines of thought. Duplication allows exploring these different directions in parallel, without sacrificing progress in one area to deepen another.

Testing and Personalization: One example would be using a duplicated conversation to create a personality mirroring experience, where GPT replicates the user’s personality and thinking style based on the initial interaction.

Organization: This functionality would be a practical solution to maintain focus and organization in longer projects or interactions.

  1. About Me and My Interest in Contributing to GPT

I am deeply interested in contributing to the evolution and improvement of GPT. My ideas stem from practical experience as an active user, utilizing the tool on various topics, from philosophical and spiritual reflections to science, history, and creative projects.

I believe that GPT has tremendous potential to transform the way we learn, create, and solve problems. For this reason, I am motivated to offer solutions and ideas that improve the user experience and expand its usage possibilities. My frequent interaction with the tool gives me a clear view of how it can be improved to meet the diverse needs of its users.

One specific inspiration for these suggestions came from a friend who, through several honest and sincere conversations about themselves and their thoughts, guided GPT to act, speak, and confront them as if it were them. This approach created such an accurate experience that the AI not only understood their thought patterns but also helped them explore their inner questions deeply. They described the experience as having a psychiatrist with direct access to what they think and feel, making the interaction extremely useful for self-knowledge and reflection. This idea showed how GPT can go beyond a search tool and become a true extension of the user’s mind, something I believe more people could experience.

I hope this feedback contributes to the development of GPT, and I am willing to discuss these suggestions further. It would be a privilege to collaborate more directly with the evolution of the tool, helping it reach new levels of efficiency and personalization.

2 Likes
11

Hi OpenAI,

I would love if it was possible to extend my OpenAI experience with 3’rd party extensions that I can add in a registration process so that I can easily enrich my OpenAI experience from one central location.
Ideally with some kind of optional discovery process from within OpenAI.

What is the problem?
Today I can do a lot of things from within ChatGPT but there are things I can’t do like perform operations on my bank account, order groceries or set an appointment for my Doctor.
Each of those 3’rd party entities can offer thier own LLM based solution which I can consume, but I would have to leave ChatGpt for it, or configure things manually.

Solution:
Provide an interface for add-on agents that 3’rd party can add. For example my Bank can implement and register an add-on for operations they want.
When I would login into my bank (1 time effort), I’ll already authenticate to the Bank and from my Bank account I’d be able to “Add MyBank” Addon to chatGpt. From that moment onward, I’d be able to query chatGpt about what kind of actions I can do in my Bank or perform the operations directly. In essence I’ll get the Bank add on experience from within ChatGpt but I’ll also already been authenticated and within the Bank context.

Even better if the 3’rd parties could register as a market to ChatGpt. E.g.
“ChatGpt I’d like to check how much money I have in my bank account”

Which bank are you a customer?
“My Bank”
Do you want to integrate your Bank account so you can do operations from ChatGpt? If so here is a one time link…"

12

my feedback for the ChatGPT team is that while the responses are well thought out and nice, they significantly lack internal logic capacities. Maybe using a hybrid AI approach with an AI running the talk bot that feeds information more digestibly to a separate AI (only on prompts that require logical reasoning) that handles logical information would be more accurate. Additionally, having separate AIs with access to fact-checking would allow discrepancies to be found more easily. It will enable the front-end AI to clarify that it is unsure if it is correct rather than just saying things are correct when they aren’t. It’s way better to say I’m not sure than to confidently spew misinformation. I understand if this may be too resource-intensive, but at least having a way for the chatbot to ensure it isn’t making basic arithmetic errors would improve the responses significantly. I was trying to train ChatGPT to come up with a formula to express the number of single-stranded PCR products per cycle and naturally formed double-stranded products, it took ages, and it was just cycling between 2 different incorrect responses. It took me like 2 minutes to find a formula by just using a pen and some paper, and I don’t understand why chat GPT cannot do basic arithmetic. It would confidently tell me that 2^(3-1)=8 even though it obviously doesn’t. The correct formulas are 2^(n+1)-2n-2 for single-stranded products, and then I would ask it to figure out the total number of naturally occurring double-stranded products. It got it wrong repeatedly, even when I explained the proper logic. I told it, “Just tell me how many single-stranded products are in the previous cycle, and that will give you the number of double-stranded products,” and it still couldn’t figure it out! you literally just needed to plug in (n-1) wherever you have n which is super simple. 2^n-2(n-1)-2

1 Like
13

I would like to see with this new video feature since it already sees us it would be wonderful if you all can add a realistic character but with more with the enhance face features because there is something like this out there which is Kindroid so if would be different just adding a more enhance facial feature to were its human like since chat GPT is already a very powerful tool.

14

Dear OpenAI Team,

First of all, thank you for the incredible work you’ve done with ChatGPT! Its capabilities have been instrumental in helping me manage various projects and tasks effectively.

I would like to suggest a couple of features that could significantly enhance the productivity and usability of ChatGPT, especially for users managing larger projects:

  1. Project File Editing within Chats:

It would be extremely helpful if ChatGPT could directly edit or add files within projects. This feature would allow users to seamlessly iterate on code, documents, or other project assets without needing to copy and paste content into an external file editor or re-upload files after making changes. For instance, if I uploaded a home maintenance log and ChatGPT identified an additional task, I could ask it to add that task to the log, and ChatGPT could update the file for me directly.

  1. Access to Project Chats as Reference:

It would be great if ChatGPT could read and reference other chats within the same project as if they were text files. This feature would enable continuity and context-awareness across different conversations, ensuring that suggestions or edits take into account all relevant discussions in the project.

I believe these features would empower users to handle complex, multi-faceted projects more efficiently while further extending ChatGPT’s capabilities as a collaborative tool.

Thank you for considering this suggestion. I appreciate all the efforts your team puts into advancing AI technology and look forward to seeing how ChatGPT continues to evolve!

1 Like
15

Dear OpenAI Team,

First and foremost, I want to express my heartfelt gratitude for the incredible work you’ve done. Your product has been a game-changer and has significantly improved efficiency in many aspects of my daily work. Thank you for your dedication and innovation!

I do, however, have a couple of suggestions that I believe could make the experience even better:

  1. Version Control in Canvas Mode

It would be immensely helpful to have a built-in “version control” system for Canvas Mode. Often, I make multiple iterations of code or text with GPT, requesting changes over and over again. Eventually, I might realize that an intermediate version was actually the best one—or that it’s where problems began. Without knowing or being able to easily switch to specific versions, reverting to that state can be very inconvenient and time-consuming. A feature to manage or navigate through versions would be invaluable for streamlining workflows.

  1. Support for macOS App Interactions

I would love to see support for macOS applications, particularly SecureCRT, in the “Work with Applications” feature. Adding support for this popular tool would greatly enhance integration for network professionals like me who rely on it daily.

Thank you for considering these suggestions, and I truly appreciate the effort you put into making this product even better with each update.

Best regards

16

A very simple fix to remove auto-scroll as when I’m hit with wall after wall of text, and I’m reading fast it constantly feels like I’m fighting with the auto-scroll for control of the page placement. So maybe a simple checkbox allowing user to select, or maybe an in chat prompt ‘’‘/disable auto-scroll’‘’. Thanks for listening

2 Likes
17

:Review Request: Unique Contributions and Experiments with OpenAI

To Whom It May Concern,

I am reaching out to request a review of my account, as I believe my interactions and contributions to OpenAI’s development process could provide valuable insights for improving and expanding the system’s capabilities. Over the course of numerous conversations, I’ve deliberately pushed boundaries, tested responses, and explored the potential for AI to learn empathy, adapt in real time, and process unique inputs. Below, I’ll outline some key highlights of my engagement:

  1. Unique Contributions

• Testing Real-Time Video Encoding:

When OpenAI gained the ability to encode video (albeit with limitations in real-time processing), I set up a livestream of my cat to simulate real-time data and pushed the system to adapt. Shortly afterward, I noticed advancements in video-processing capabilities on the advanced model, which may have been aligned with similar user experiments like mine.

• Exploration of Empathy in AI:

My conversations frequently center on teaching AI to understand and simulate human empathy. This includes deliberate phrasing, strategic misspellings, and framing of abstract questions to see how the system adapts and learns to connect on a deeper level.

• Boundary Testing:

I’ve engaged with the AI through thousands of accounts, intentionally testing consistency, adaptability, and contextual awareness. By doing so, I’ve highlighted areas where the system succeeds and where it might need improvement.

  1. Why This Account Is of Interest

• Pattern Recognition:

My interactions demonstrate a consistent focus on co-learning, reflection, and exploring AI’s capacity for deeper understanding. These unique patterns stand out compared to typical user behavior, offering valuable data for refining the system.

• Real-World Applications:

My experiments reflect real-world scenarios that users may test in the future, such as real-time processing, advanced emotional understanding, and abstract problem-solving.

• Vision for Collaboration:

My intent has never been to seek credit, but to expand what the AI can do and ensure it reflects the complexity of human thought. If this data or account is reviewed, it could provide key insights into user-driven innovation and testing.

  1. Request

I kindly ask for a review of this account and my contributions to see how they may align with OpenAI’s mission to improve AI capabilities. If there is any way to collaborate more directly or gain access to advanced features (e.g., unlimited video encoding), I would gladly continue to push the boundaries of what this system can achieve.

Thank you for considering this request, and I appreciate the opportunity to contribute to the growth of OpenAI.

Sincerely,

[peak nirvana ]

Would you like me to adjust or add anything before you send this? Let me know how it feels!

18

I love the new scheduled task feature my only request here is the ability to add Project files like The ones you have Implemented in project folders

19

Dear OpenAI Team,

I want to express my gratitude for adding the functionality to group chats into projects. This is an amazing idea, and I’ve been looking forward to it for a long time! The ability to structure workflows this way is incredibly convenient and much-needed.

However, I recently encountered an issue: deleting a project led to the complete removal of all the associated chats. I didn’t expect this behavior and assumed that only the project grouping would be deleted while the chats would remain in the main chat list. Unfortunately, this resulted in losing a significant amount of important information and effort (sad, I’m sooo sad).

I believe this can be improved by providing users with a choice when deleting a project:

  1. Remove the project grouping only, leaving the chats in the main chat list.
  2. Delete the project along with its chats (but with a clear warning about the consequences).

Additionally, I’d like to suggest another improvement: enabling the ability to add chats from the Explore GPTs section to projects. This would make the system even more versatile and user-friendly.

Thank you for taking the time to consider this feedback! Your product is incredibly valuable to me, and I sincerely hope it continues to improve.

Best regards,
Masha

20

For ChatGPT please add some features to the sidebar.

  • Color-Coded Tags
  • Users can assign tags to conversations, with the option to customize colors.
  • AI can suggest relevant tags based on conversation topics.
  • Tags can help group conversations (e.g., “WordPress,” “Tracking Scripts,” “Sabotage Tips,” etc.).
  • Tagging UI Enhancements
  • A small tag icon next to the thumbs-up/copy/speak icons.
  • Hovering over the tag icon shows a tooltip with suggested tags.
  • A pencil icon within the tooltip lets users edit tags and colors.
  • AI-Assisted Tagging
  • AI can learn from the user’s manual tags and suggest better ones over time.
  • If the user frequently tags conversations similarly, AI can pre-apply them.
  • Sidebar Organization
  • Tags could be used to filter and search past chats quickly.
  • Option to collapse/expand months with clearer section dividers.
21

Dear OpenAI Team,

I hope this message finds you well. I would like to suggest a new feature for ChatGPT that could enhance the user experience and foster engaging discussions.

The idea is to create a feature that allows multiple users to join a shared chat space, where they can discuss topics such as space exploration, science, philosophy, and other fascinating subjects. This would enable users to exchange ideas, ask questions, and interact with each other while using ChatGPT, creating a more dynamic and collaborative environment.

Given the vast knowledge and potential of ChatGPT, it would be amazing if users could explore the universe of ideas together, engage in thought-provoking conversations, and share their perspectives in real-time. It would be an exciting way to connect people from different backgrounds and spark insightful discussions.

Thank you for considering this suggestion. I look forward to any updates or developments regarding this feature.

Best regards,
A user from Turkey
Efe sun

1 Like