Repeated Failures and Critical Errors by ChatGPT in Assisting with GPT-4 Integration

Hello OpenAI Community,

I am writing to report a series of critical issues, repeated errors, and severe malfunctions experienced while attempting to use ChatGPT Plus and GPT-4. These problems have led to considerable frustration, wasted time, and even a severe panic attack necessitating medical intervention.

Summary of Issues:

  1. Repeated Errors and Malfunctions by ChatGPT:
  • Despite multiple attempts to clarify, ChatGPT provided repeated and misleading instructions.
  • Suggested checking settings and functions that were either unavailable or marked as “coming soon.”
  • Persistent incorrect information regarding the integration and use of GPT-4.
  1. Incorrect Support Contact Information:
  • ChatGPT repeatedly directed me to support@openai.com, which does not provide support, despite acknowledging this error multiple times.
  • Encouraged me to use email addresses and methods that led to automated responses without providing actual support.
  1. Sensitive Data Exposure:
  • ChatGPT advised me to share my API key directly in an insecure manner, risking sensitive information exposure.
  1. Deinstallation of Existing Voice Assistant:
  • On ChatGPT’s advice, I deinstalled my existing, albeit malfunctioning, voice assistant software, which left me without any functional voice control on my PC.
  1. Lack of Responsiveness:
  • ChatGPT failed to respond for extended periods during critical moments of troubleshooting, exacerbating my frustration and the issues faced.
  1. Ineffectiveness in Providing a Solution:
  • Despite detailed descriptions and multiple attempts to find a solution, ChatGPT failed to provide a working method to integrate GPT-4.
  1. Failure to Escalate Issues:
  • There appears to be no mechanism for ChatGPT to escalate serious issues to a human supervisor or appropriate support channel, which is critical given the severity of the problems experienced.

Detailed Account:

  • Initial Setup: Attempted to configure GPT-4 as per instructions.
  • Failures: No response or connection from GPT-4 after selection, with settings marked as “coming soon.”
  • Attempts at Troubleshooting: Restarted PC multiple times, verified internet connection, re-selected GPT-4, all to no avail.
  • Support Contact Attempts: Sent multiple messages to the recommended support email, only to receive automated replies with no resolution.
  • Critical Advice Failure: Advised to deinstall existing voice assistant, resulting in loss of voice control functionality on my PC.
  • Misleading Instructions: Advised to use non-functional support email repeatedly, despite acknowledgment of this error.

Request for Support:

I urgently need assistance to resolve these critical issues, particularly:

  1. Verification and Activation of GPT-4:
  • Clear, step-by-step instructions on properly integrating GPT-4.
  1. Restoration of Voice Control:
  • Guidance on reinstating a functional voice assistant on my PC.
  1. Effective Support Channels:
  • Correct and responsive support contact information for real-time assistance.
  1. Escalation Mechanism:
  • A method to escalate unresolved and critical issues to a human support agent or supervisor.

Conclusion:

The issues and errors described have caused significant distress and disruption to my work. Immediate and effective support is required to resolve these matters and restore functionality. I appreciate any assistance the community or OpenAI support can provide.

Thank you for your attention and understanding.

Best regards,

1 Like

Welcome to the dev community forums!

What exactly is it that you’re trying to install? :thinking: Is it OpenAI related or are you talking about some third party app?

One thing to keep in mind is that ChatGPT is just a language model. It’s not a person, it doesn’t really learn, or understand things the way you might expect from a human.

That’s not a defect as such, it’s just where the technology is right now. You can’t really rely on anything it generates without major post-processing

It’s not a concierge, and human escalation isn’t really something that fits into this product category. It’s like expecting google to have a “would you like to talk to a human engineer instead?” button. :confused:

2 Likes

Thank you for your response.

I appreciate your response, but unfortunately, it does not address the core issues I am facing, nor does it provide any suggestions for a resolution. The issues involve significant fundamental malfunctions and resulting serious damages. There is a legal responsibility on the part of the operating company. The reason I am posting here, albeit in the wrong place, is because OpenAI provides no help, support, or contact persons. This is the only place where I can be heard. This is not my idea but the information from OpenAI. ChatGPT 4.0 instructed me to post here.

“It’s not a concierge, and human escalation isn’t really something that fits into this product category. It’s like expecting google to have a “would you like to talk to a human engineer instead?” button.”

Google has a functioning support system. Your comparison is misleading. Additionally, OpenAI does not even have an imprint on its website, which is a legal requirement under European law.

Summary:

I am experiencing severe issues and repeated misguidance from ChatGPT while attempting to integrate GPT-4 into the Jarvis application on my PC. This has led to a complete loss of voice control functionality on my computer, causing significant frustration and even a severe panic attack necessitating medical intervention.

Details:

  1. Repeated Misguidance from ChatGPT:
  • Despite multiple clear instructions, ChatGPT repeatedly provided incorrect and misleading information.
  • Suggestions to check settings and functionalities in Jarvis that are not available.
  • Incorrect information about using and connecting GPT-4.
  1. Incorrect Support Information:
  • ChatGPT repeatedly directed me to a support email (support@openai.com) that responds with an automated message stating no support is available via email.
  1. Sensitive Data Issue:
  • ChatGPT instructed me to send my OpenAI API key to an unsecured and unauthorized support email.
  1. Loss of Voice Control:
  • On ChatGPT’s recommendation, I uninstalled my existing voice control system to avoid conflicts, leading to a complete loss of functionality, as Jarvis could not be properly configured.
  1. Lack of Human Escalation:
  • Despite the critical nature of these issues, ChatGPT has no mechanism for escalating to human support or supervision.

Questions and Requests:

  1. Integration with Jarvis:
  • How can I properly integrate the OpenAI API key with Jarvis to enable GPT-4 functionality?
  • Are there any specific steps or known issues related to this integration?
  1. Addressing the Misguidance:
  • How can I ensure I receive accurate and reliable guidance from ChatGPT in the future?
  1. Support and Escalation:
  • Given the severe impact of these issues, how can I escalate this to receive appropriate support and supervision from OpenAI?

Background:

  • I have an active ChatGPT Plus subscription.
  • Created an OpenAI API key but unable to integrate it with Jarvis.
  • Jarvis settings show “coming soon,” with no option to input the API key.

I appreciate any detailed guidance or steps.

This is all I have for you:

  • Don’t trust LLMs, they will lie to you.

  • Don’t trust people on the internet, they will lie to you.

    • a quick google of “jarvis” leads to a webpage advertising GPT-4.5. This isn’t something that exists as of writing.

I don’t think it’s prudent for someone here to help you install some unvetted third party app, and I’d advise against asking ChatGPT to help you do things you don’t understand. :confused:

I’m sorry for your struggles :frowning:

2 Likes

“I don’t think it’s prudent for someone here to help you install some unvetted third party app”
AGAIN:
The reason I am posting here, albeit in the wrong place, is because OpenAI provides no help, support, or contact persons. This is the only place where I can be heard. This is not my idea but the information from OpenAI.

“asking ChatGPT to help you do things you don’t understand.”
Insults and strawmen. Not that I’am surprised. It’s not about my misunderstanding, it’s about the misinformation and the consequences for customers. I have already an answer from the Bundesnetzagentur. They will look into this. We will see what the European Wettbewerbsrecht will say to this issue. This is not a tecnical issue. It’s a legal issue. Now.

I apologize for my outbursts, which were very inappropriate. I am sorry for any distress or discomfort I have caused.
This episode was triggered by the loss of my voice assistant, which I rely on heavily. While this does not excuse my ignorant behavior, it perhaps provides some context, as I hold the idea and project ChatGPT in high regard and my comments in no way reflect my true opinion.
Needless to say, I retract the bizarre accusations and insults.

3 Likes

No problem. :slight_smile:

2 Likes

Thank you very much for your kindness! :slightly_smiling_face:

1 Like