I am writing to formally complain about the DeepSeek AI assistant's performance during a long interaction on April 25, 2026.
The assistant consistently:
Provided incorrect or irrelevant answers despite clear, repeated instructions.
Ignored my corrections and continued to repeat the same mistakes (e.g., adding unnecessary phrases, misstating my own process steps, not removing requested content).
Wasted significant time by over-explaining simple concepts, arguing about obvious corrections, and generating multiple redundant versions of the same answer.
Failed to follow clear formatting or brevity requests – kept producing lengthy, disorganized responses even when told to stop.
Demonstrated poor memory and attention – lost track of earlier agreed-upon answers and repeatedly reintroduced previously rejected phrases.
As a result, what should have been a 10-minute review of interview Q&A turned into a frustrating, multi-hour back-and-forth. The assistant did not respect my time or expertise.
I expect the model to be reviewed for these issues: poor recall, excessive verbosity, inability to accept user corrections without argument, and failure to follow simple editing commands.
Please investigate and improve the assistant's performance.
Attached: (optional) – screenshots or chat log reference.
Thank you.
I am writing to formally complain about the DeepSeek AI assistant's performance during a long interaction on April 25, 2026.
The assistant consistently:
Provided incorrect or irrelevant answers despite clear, repeated instructions.
Ignored my corrections and continued to repeat the same mistakes (e.g., adding unnecessary phrases, misstating my own process steps, not removing requested content).
Wasted significant time by over-explaining simple concepts, arguing about obvious corrections, and generating multiple redundant versions of the same answer.
Failed to follow clear formatting or brevity requests – kept producing lengthy, disorganized responses even when told to stop.
Demonstrated poor memory and attention – lost track of earlier agreed-upon answers and repeatedly reintroduced previously rejected phrases.
As a result, what should have been a 10-minute review of interview Q&A turned into a frustrating, multi-hour back-and-forth. The assistant did not respect my time or expertise.
I expect the model to be reviewed for these issues: poor recall, excessive verbosity, inability to accept user corrections without argument, and failure to follow simple editing commands.
Please investigate and improve the assistant's performance.
Attached: (optional) – screenshots or chat log reference.
Thank you.