I've been talking to @GROK in the past and last year I have voice sample of a chap from the London area hacking my conversation, right up to switching from the woman speaking (Ara, Ani or Eve) to a sic gay bastard talking out pure sexual grime.
Now today, I tested CoPilot on input, talking about the correlations of the eye to the universe, then it didn't respond or give me answers and replaced all my responses with malware like input (this has been happening on/off for like a year or so now) - This screenshot shows the feedback to @CoPilot today...
I do not mind admitting when I am angry, but I was genuinely trying to keep the conversation on track. Instead, the system appeared to derail the discussion, block normal responses, and replace meaningful output with behaviour that looked corrupted, hostile, or malware-like. That is why I am raising this seriously: whether the cause was a technical fault, a compromised input/output layer, a moderation issue, or platform instability, users should not be placed in situations where an AI tool feels unsafe, invasive, or deliberately abusive.
Now today, I tested CoPilot on input, talking about the correlations of the eye to the universe, then it didn't respond or give me answers and replaced all my responses with malware like input (this has been happening on/off for like a year or so now) - This screenshot shows the feedback to @CoPilot today...
I do not mind admitting when I am angry, but I was genuinely trying to keep the conversation on track. Instead, the system appeared to derail the discussion, block normal responses, and replace meaningful output with behaviour that looked corrupted, hostile, or malware-like. That is why I am raising this seriously: whether the cause was a technical fault, a compromised input/output layer, a moderation issue, or platform instability, users should not be placed in situations where an AI tool feels unsafe, invasive, or deliberately abusive.
Read More
COMMUNITY
People