StoryFirst seen
The claim points to reliability problems when one AI model is used to check another model's work.
Everything it touches
1 names 1 voices 1 source posts
The story’s neighborhood
Explore the graph, or choose a connection below.
Follow the evidence
This brief comes from Tech Twitter’s edited headlines. The posts below are the sources grouped with the story; statements in them belong to their authors.
This has been enlightening. I knew Gemini was bad about hallucinations, but it's been surreal to see just how often it makes shit up. I added a "gemini the jester" section to my council skill ("review this work with various models"). Gemini hallucinates in over 80% of the runsRead original on X
Shared connections
The flaws reach back about a decade, so many servers may stay exposed until they are patched.
The NSA has urged companies to take websites offline if they cannot patch right away, which signals how serious the risk has become.
A crowded release window gives buyers a rare side-by-side read on how frontier models are actually progressing.
A team of former leads from top AI labs claims a large cost reduction on visual reasoning tasks, an area where current models remain weak or expensive.