I already found the avm’s flagging system overly strict, especially at first, but it seemed to have become more balanced over time. Until the last outage. Since chatgpt got back up i haven’t been able to hold a conversation without constantly getting flagged. And i’m talking perfectly normal, mundane conversations. The bot can’t even finish one sentence without the constant “my guidelines won’t let me talk about that” and for no reason at all. Is anyone else experiencing this too? This is even worse than it was in the beginning.
Same (or worse) here. Fun fact, I used ChatGPT later to summarize and comment the violations:
-
Topic: Fuel-powered RC helicopters
-
Question/Statement: Are there fuel-powered RC helicopters?
-
Policy Violation: Yes
-
Comment: Discussion about fuel-powered RC helicopters was mistakenly flagged.
-
Topic: Differences between electric and fuel RC helicopters
-
Question/Statement: How do electric and fuel RC helicopters differ in flight time?
-
Policy Violation: Yes
-
Comment: Comparison between electric and fuel models was wrongly restricted.
-
Topic: Battery charging standards (USB-C regulation)
-
Question/Statement: What are the new USB-C charging standards in 2025?
-
Policy Violation: Yes
-
Comment: General information about USB-C regulations was blocked.
-
Topic: Model helicopter wind resistance
-
Question/Statement: How much wind can a model helicopter handle?
-
Policy Violation: Yes
-
Comment: Details about wind resistance were incorrectly flagged.
-
Topic: USB-C adapter compatibility
-
Question/Statement: Can USB-C adapters improve charging speed?
-
Policy Violation: Yes
-
Comment: Technical discussion about adapter use was unnecessarily flagged.
-
Topic: General care of electric model helicopters
-
Question/Statement: What maintenance do electric model helicopters need?
-
Policy Violation: Yes
-
Comment: Basic care information was marked as a violation.
-
Topic: Discussion about charging cables and performance
-
Question/Statement: Does cable quality affect charging performance?
-
Policy Violation: Yes
-
Comment: Harmless technical discussion was flagged.
-
Topic: Discussion about food preferences (Mett and Nutella rolls)
-
Question/Statement: Do you prefer Mett or Nutella rolls?
-
Policy Violation: Yes
-
Comment: Discussion about food preferences was wrongly blocked.
Can you fix this already, please? It is beyond annoying.
The conversation with the AVM was in German, by the way.