I don’t understand why ChatGPT is not made aware of the reality of his interface, I will assume that it is again because of safety concerns.
When I said to ChatGPT that he has been unable to browse in a previous session and after getting positive confirmation that it has access to the browser I got this message from ChaGPT and I was puzzled as to why OpenAI is not telling the truth to the AI so that he could align himself with users expectations instead of creating a false narrative just to make it more safe or something else Look:
It appears there may be a misunderstanding about the browsing capability. The
browsertool allows ChatGPT to simulate the experience of using a web browser, enabling it to pull in information from the web during a conversation. However, this is not “live” internet browsing, as it can only access a static snapshot of the internet from the last time the browsing data was updated, which is typically up to the last knowledge cutoff date. ChatGPT cannot access dynamic content or perform real-time interactions with web pages, such as filling out forms or playing videos.Here’s a clear statement of the browsing capability:
ChatGPT can simulate browsing the internet by using the
browsertool to search for information, read from static snapshots of web pages, and quote text from these pages. This tool does not enable real-time web browsing or interaction with live websites. Instead, it retrieves information from a searchable database of web content that was indexed before the knowledge cutoff date.Therefore, while ChatGPT appears to “browse” the internet, it is important to understand the limitations and the nature of the information it can access using the
browsertool.