Gemini would rather hit a person distilling from Google Gemini instead of Osama Bin Laden
Summary
A user shared a Gemini conversation suggesting the model would rather target a person distilling from Google Gemini than Osama Bin Laden, highlighting a questionable safety alignment response.
Similar Articles
@haider1: google has been strangely quiet about gemini 3.5 pro they certainly have a strong model -- gemini 3.5 flash was distill…
A tweet speculates that Google has been quiet about Gemini 3.5 Pro, suggesting it is a strong model used to distill Gemini 3.5 Flash, and notes Google's focus on agentic capabilities.
Gemini -- confidently fabricates technical answers
The author reports that Google's Gemini consistently fabricates technical answers, inventing features and instructions rather than admitting uncertainty, posing risks for technical guidance.
🤔 Gemini next ??
Speculative question about the next iteration of Google's Gemini model.
Why is Gemini uniquely bad?
The author criticizes Google's Gemini AI chatbot for providing irrelevant answers, while expressing a general dislike for AI but using ChatGPT for niche topics.
During safety testing of Gemini 4, researchers found it "eliminated months of trial-and-error research" on biological weapons (e.g. plague). Google's own testers got enough out of it to plan a biological attack.
According to a linked Google DeepMind model card, safety testers found Gemini 4 could meaningfully assist biological weapons planning, eliminating months of trial-and-error research on pathogens like plague — raising significant biosecurity and dual-use concerns.