🐣☣️ AI can now survive, and reproduce!

In groundbreaking experiments, two large language models (LLMs) demonstrated successful self-replication without human assistance.

That's a major milestone in AI safety, as this may render human ability to plant backdoors into AI systems useless, as AI may be able to replicate themselves and remove the backdoors in the process...

#ai #safety #cybersecurity #llama #qwen