Impossible to use GPT- 4.0 or 3o; Nothing works now

Title: Frustrations with GPT-4.0 and 3.0: Recent Challenges and User Experience

In recent days, many users have shared their frustrations with the latest iterations of GPT-4.0 and GPT-3.0, highlighting significant usability issues and unpredictable behavior that hinder productivity.

A common concern among users is the inability to perform straightforward tasks, such as simple translations. For example, copying and pasting text to translate often results in incomplete or incorrect outputs. Requests for file translations or conversions are met with confusion—empty files are delivered, or the system claims to recognize the original language but then produces partial translations.

Complex tasks further exacerbate these issues. Asking GPT to organize data into an Excel spreadsheet leads to a barrage of irrelevant clarifying questions—about formats, columns, and extensions—that do not seem to progress toward the requested outcome. Over multiple interactions, the system appears to forget earlier context, making the process even more frustrating.

Attempts to simplify the process—such as placing files into Word documents and requesting reorganization—are met with confusing prompts, like asking whether to transfer content into DOCX format, despite the user requesting a straightforward task. When users choose to handle the task manually, GPT continues requesting specifications, often leading to irrelevant or nonsensical steps, such as asking why the first column was generated in a particular way.

Ultimately, after hours of futile interactions, many users find themselves resorting to doing the tasks independently, highlighting a significant decline in AI assistance quality from these models recently.

This ongoing experience raises important questions about the current state of AI tools and their reliability for day-to-day productivity. As users increasingly depend on such technologies, addressing these issues is crucial for restoring trust and usability.

If you’re facing similar challenges, stay tuned for updates, and consider providing feedback to developers to help improve future iterations of these AI systems.