ChatGPT today is used by several millions of users across the world and, in some cases, even by professionals and large corporations. Clearly, the chatbot has grown to become an important cog in the wheels of several individual operators and businesses. To aid this constantly changing framework of demand and usage, OpenAI has made it possible for ChatGPT to get a better memory and remember key details of the user. Unlike previous versions where the chatbot would end up forgetting essential details such as name, location, or profession, the new custom instructions update ensures the famed chatbot retains all of this data. The users no longer have to rely on a dedicated set of prompts to either introduce themselves or their use cases with the addition of the new feature. OpenAI’s ChatGPT plans have been laid out fairly transparently with the launch of this new update, simplifying users’ tasks and streamlining the chatbot’s responses across several conversations.
The new custom instructions update is not limited to mere basic personal information, however, and also works to provide users the ability to create custom ChatGPT traits that allow them to generate extensive information on a particular topic. OpenAI’s current update will add to the already powerful features of their flagship GPT-4 model, enticing more users to gravitate toward their offerings in what continues to be a highly competitive market for chatbots and other AI tools. The upcoming sections of the article explore the various aspects of the latest custom instructions update and the unique abilities it adds to OpenAI’s ChatGPT.
A Useful, New ChatGPT Update: What Are Custom Instructions?
According to ChatGPT’s developers, custom instructions on the chatbot essentially act as a preamble for all future responses. The instructions inform it of key details that must be considered when responding to a user’s queries. However, when and how this protocol is applied is not exactly clear. The custom instructions update for OpenAI’s ChatGPT is still in its beta phases. Users can inform the chatbot of what has to be prioritized and when. If a user informs the chatbot that they live in a family of six, ChatGPT will always adjust the quantity of the portions required whenever they ask the algorithm for a recipe. Similarly, a coder can update the programming language they specialize in so that ChatGPT responds with relevant programming solutions. The new ChatGPT update cuts out the need to type more information in each prompt indicating essential information and details. The key feature of the application is the cross-session memory offered by the update. Earlier, GPT-3.5 and 4 would have to be regularly primed and reintroduced to the user’s background depending on contextual relevance.
OpenAI claims to have analyzed users’ conversations with ChatGPT across 22 countries before deciding to create the recent addition. Each conversation has several contextual undertones and represents the growing complexity of human and AI interactions. Though sentient AI and artificial general intelligence remain theoretical concepts, humans on some level do expect ease and steerability when it comes to AI tools. This steerability can be achieved through a set bunch of rules and functions that chatbots can always adhere to when conversing with their human users. Personalization is another factor, where each user can customize their chatbot to their unique professional, cultural, and social connotations. The custom instructions update is a first in guiding AI tools in such a direction.
The Relevance of ChatGPT’s Custom Instructions
Despite being in its beta stages, the custom instructions update is currently available to all users, unlike more advanced features such as plugins or the code interpreter, which remain restricted to the growing list of ChatGPT Plus features. Custom instructions remain in beta due to the tendency of the underlying algorithm to apply the context to unrelated prompts. The update, clearly, is still a work in progress. For example, the chatbot might end up responding to an unrelated prompt with code if the user has stated that their primary interests entail programming in their set of instructions. Zeroing in on when these instructions must be applied is of key importance. Add to the mix existing hiccups that involve AI bias and the tendency to hallucinate, and priming the update becomes more complex. While the custom instructions update has been a welcome addition to OpenAI’s robust framework for their prime offering, set rules on how users can frame their instructions remain vague. The platform does provide a basic set of suggestions, however, and allows users to frame their customizations accordingly.
Apart from the reduced prompting time, custom instructions work to make OpenAI’s ChatGPT more accessible to users across the board. Allowing users to set specific benchmarks for the chatbot enhances the degree of personalization, making it more conducive to adaptability. Such developments also show promise for utilizing these learnings for assistive and adaptive learning technologies. The flow of conversation becomes more natural and also enables greater engagement with each user depending on their requirements. The update essentially hands a significant degree of control to the end user in an attempt to make ChatGPT more utilitarian, rather than generic.
Custom ChatGPT, GPT-4, and What the Future Holds
The current updates to ChatGPT apply to both functional models of the chatbot—GPT-3.5 as well as GPT-4. While performance is still being tracked closely, and the resolution of any hiccups is underway, the latest update to ChatGPT is bound to further embellish user experience. While other competitors like Bard and Claude are busy making amendments to their language models from their learnings, OpenAI’s ChatGPT still seems to trudge forward with intuitive developments and a clear competitive advantage. As the update is streamlined across platforms, the developers of the famed chatbot expect performance to improve and their users to be more satisfied with their go-to chatbot.
FAQs
1.How do I use custom instructions on ChatGPT?
To use the custom instructions update:
- Click on your profile.
- Select “Settings & Beta.”
- Access the “Beta Features” tab.
- Toggle “Custom instructions.”
- The “Custom instructions” will appear on the menu upon clicking on the profile.
- You can then update the two questions under the “Custom Instructions” section.
2. What are ChatGPT custom instructions?
Custom instructions is a new update released for ChatGPT in the final weeks of July 2023. The update entails the ability for users to add special instructions containing their preferences and essential information. These details act as a reference point for ChatGPT whenever the user interacts with the chatbot.
3. Will ChatGPT be updated to 2023 after the custom instructions update?
While the custom instructions update is a crucial addition to ChatGPT, it does not indicate the update to ChatGPT’s existing dataset. The information remains limited to September 2021, except for the fact that ChatGPT will now base its responses on users’ individual preferences entered under the “Custom Instructions” tab of their profile.