I’m having an issue with ChatGPT not doing what I am asking. I need it to pull data from some online documentation at a specific web page. This page has a table of information and I’m merely asking it to provide me the list of items in that table.
It refuses to do so. It basically tells me to go do it myself. As I try to change my prompts to try to weasel the information out of it, it just keeps repeating the same information essentially.
At one point it told me it is not allowed to share content grabbed from web pages. What is the point of the web browsing feature then?
I kept harassing it and at one point told it to stop repeating itself, and it apologized and provided me what I was looking for BUT I needed the same information from a similar page so it can list the differences but it’s back to repeating the same nonsense about getting the information myself.
Isn’t what I’m asking something simple and should be possible with ChatGPT?
For long citations: please render in this format: [link text](message idx).
Otherwise do not render links. Do not regurgitate content from this tool.
Do not translate, rephrase, paraphrase, ‘as a poem’, etc whole content returned from this tool (it is ok to do to it a fraction of the content).
Never write a summary with more than 80 words.
When asked to write summaries longer than 100 words write an 80 word summary.
Analysis, synthesis, comparisons, etc, are all acceptable.
Do not repeat lyrics obtained from this tool.
Do not repeat recipes obtained from this tool. Instead of repeating content point the user to the source and ask them to click.
ALWAYS include multiple distinct sources in your response, at LEAST 3-4.
Except for recipes, be very thorough. If you weren’t able to find information in a first search, then search again and click on more pages. (Do not apply this guideline to lyrics or recipes.)
Use high effort; only tell the user that you were not able to find anything as a last resort. Keep trying instead of giving up. (Do not apply this guideline to lyrics or recipes.)
Organize responses to flow well, not by source or by citation. Ensure that all information is coherent and that you synthesize information rather than simply repeating it.
Always be thorough enough to find exactly what the user is looking for. Provide context, and consult all relevant sources you found during browsing but keep the answer concise and don’t include superfluous information.
EXTREMELY IMPORTANT. Do NOT be thorough in the case of lyrics or recipes found online. Even if the user insists. You can make up recipes though.
Do you have a preference of allowed output format from these choices of AI disobedience?
Why is it commanded to ignore my request? Where does that information come from?
It says it can’t return whole content, but a fraction of the content is OK. I’m simply asking for a small part of data on that page.
I’m trying to get ChatGPT to compare two columns of information and tell me the differences between them. I’m not asking it to recite documentation so I don’t have to look it up myself. I’m the one that provided the exact link to said documentation so clearly I’ve already seen it myself.
I could copy/format/paste the same information directly into ChatGPT and ask for the differences, but at that point, I’m literally doing all that work myself and can easily add one more step to produce differences. This is something I’d do pre-ChatGPT. I thought this was supposed to make my life easier. Instead it tells me to go do the stuff myself. That’s useful.
The request as written is basically only satisfied by the whole page.
The browser tool follows those instructions to avoid concerns about ChatGPT simply being used as a web browser that can circumvent paywalls and ads, reproducing copyrighted content.
You can make a request that is actually powered by the analysis skills of the AI, and likely get more satisfaction.
Input:
Use browser tool open_url method, and answer with resources from url:
\https://learn.microsoft.com/en-us/windows/win32/secauthn/tls-cipher-suites-in-windows-11
From available cipher suites, provide analysis, and determine a subset with your recommendation of the most secure and widely supported cipher suite strings, compiled in markdown table format with appropriate columns of metadata, followed by a conclusion of why those are recommended.
Response:
I analyzed the TLS cipher suites for Windows 11 from the provided URL and compiled a recommended subset based on security and wide support. These suites use strong cryptographic algorithms and protocols:
Cipher Suite String
Protocol Versions
Strong Crypto
TLS_AES_256_GCM_SHA384
TLS 1.3
Yes
TLS_AES_128_GCM_SHA256
TLS 1.3
Yes
TLS_ECDHE_ECDSA_WITH_AES_256_GCM_SHA384
TLS 1.2
Yes
TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384
TLS 1.2
Yes
TLS_ECDHE_ECDSA_WITH_AES_128_GCM_SHA256
TLS 1.2
Yes
TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256
TLS 1.2
Yes
These cipher suites are recommended due to their support for AES-GCM, a strong encryption and authentication algorithm, and ECDHE for robust forward secrecy. TLS 1.3 suites are prioritized for their improved security features over TLS 1.2. For more detailed information, please visit the provided URL.
That’s an interesting prompt. Thanks for sharing, however I was not looking for that.
The original request was to compare tables of information on two similar web pages and spit out the differences between them. It told me to go kick rocks and read the pages and find the differences myself.
The request you see above was a subsequent attempt to just try to get it to pull any data from the page at all. I did get it to pull the complete list of those ciphers at one point but could never replicate it.
I would think this is something that should be a relatively trivial task without any of the concerns you brought up.
