AIThe LLM safety argument built on incident counts and user numbers
One post cites 362 tracked LLM incidents in 2025 and 1.2 billion users to question the focus on hypothetical future harms.
TLDR
A post compares 362 tracked incidents involving large language models in 2025 with 1.2 billion users, calculating a rate of about 0.000030%. It suggests LLMs may be the “safest products in history” while questioning concern about future harms. A reply jokes, “What we should do is ban all cars,” claiming cars are much more dangerous than LLMs.
Combined views
624
2 Sources, first seen 15h ago
The LLM safety argument built on incident counts and user numbers
One post cites 362 tracked LLM incidents in 2025 and 1.2 billion users to question the focus on hypothetical future harms.
TLDR
A post compares 362 tracked incidents involving large language models in 2025 with 1.2 billion users, calculating a rate of about 0.000030%. It suggests LLMs may be the “safest products in history” while questioning concern about future harms. A reply jokes, “What we should do is ban all cars,” claiming cars are much more dangerous than LLMs.