Mastodon Feed: Post

Mastodon Feed

Boosted by baldur@toot.cafe ("Baldur Bjarnason"):
campuscodi ("Catalin Cimpanu") wrote:

LLMs that work with audio data are apparently vulnerable to attacks using human inaudible commands because nobody who put together that LLM thought to impose a limit to the human audible audio frequency spectrum

https://spectrum.ieee.org/voice-ai-audio-attacks