Everyone who has not yet registered an account with OpenAI can access its flagship conversational AI. However, the experience may not be as optimal, and all conversations will be stored in their training database unless you opt out.
The availability of this service will begin in select areas and gradually expand worldwide through word of mouth. Alternatively, you can log in at openai.com. Otherwise, you will be directed to ChatGPT, with a similar interface to that of registered users.
While you are free to engage in conversations to your heart’s content, certain features exclusive to account holders will not be accessible to you. Functions such as saving or sharing chats, using custom prompts, or other account-specific options will not be available.
Nevertheless, you have the choice to opt out of having your messages used for training purposes, which may challenge the fundamental premise of the company. Simply click on the question mark icon at the bottom right, navigate to “settings,” and activate the feature. The accompanying image is provided by OpenAI:
Image Credits: OpenAI
Notably, this free version of ChatGPT will adhere to “slightly more restrictive content policies.” What does this entail? When queried, a spokesperson provided a verbose yet somewhat vague response:
The unsigned user experience will benefit from existing safety measures, such as filtering out harmful content, already integrated into the system. Additionally, we are implementing further safeguards tailored to address other types of content that may not be suitable for users without accounts.
We have considered potential misuse scenarios for the unsigned service, drawing from our understanding of GPT-3.5 and risk assessments conducted.
The specifics of these enhanced policies remain unclear. As users flock to explore this new offering, feedback will undoubtedly shed light on the details. The director mentioned, “We value feedback and acknowledge that further refinement may be necessary.” They are likely to receive ample feedback in due course.
When asked about strategies to combat potential exploitation of the system on a large scale, especially for financial gain, a generic response was provided:
The teams overseeing misuse detection, prevention, and response have been actively involved in shaping and implementing this service and will continue to guide its development. We have also considered methods to detect and prevent abuse of the unsigned user experience.
The response lacks concrete details, suggesting a reactive rather than proactive approach to potential misuse. Predicting how individuals may exploit this opportunity appears to be a shared challenge.
The initial rollout of the free ChatGPT service will not specify which regions or demographics will have access first. The launch commences today, so stay tuned for updates on availability.
