OpenAI has introduced a new feature allowing users to access ChatGPT instantly without the need to sign in to their OpenAI account. Now, upon opening ChatGPT, users can immediately engage with the free GPT-3.5 model without any login requirements. This move mirrors Microsoft’s approach with Copilot, extending accessibility to a broader audience. However, certain functionalities will remain exclusive to users with accounts.
For instance, features like custom instructions, saving or sharing conversations, and voice interactions necessitate an account for personalization and data management. While this resembles an incognito mode for ChatGPT, it’s important to note that OpenAI utilizes conversations to enhance its models.
To address privacy concerns, OpenAI has implemented a one-click option enabling users to opt out of model training using their data. By clicking on the question button located at the bottom-right corner and accessing Settings, users can disable the toggle for “Improve the model for everyone.” This action prevents OpenAI from storing and utilizing their data for model enhancement purposes.
OpenAI has implemented additional safety measures to block potentially harmful prompts and outputs across various categories. The rollout of the ability to access ChatGPT without an account is being conducted gradually across 185 countries.
It’s noteworthy that OpenAI is providing access to the GPT-3.5 model without the necessity of signing up. However, for users seeking alternatives with performance similar to the more advanced GPT-4 model, there are options available. These include Claude 3 Sonnet, Microsoft Copilot, Mistral Chat, and others. Notably, Microsoft Copilot offers the GPT-4 Turbo model for free, which is highly advantageous.
For those not up to date with recent developments, OpenAI is currently in the process of developing an intermediate model, GPT-4.5, with speculation about the potential release of the GPT-5 model by the summer of 2024, as reported. Additionally, OpenAI has showcased a Voice Engine model capable of cloning voices using only a brief 15-second audio sample. However, due to significant associated risks, OpenAI has refrained from releasing this technology to the public.
0 Comments