New Anthropic research warns that "small AI models may be months away from being capable enough to self-replicate on commercial IoT devices" like smart ovens and TV boxes.

In testing, a 9B parameter model got inference running on 8 of 9 hardware types. Researchers ran a 26B model distributed across 15 off-brand Android TV boxes bought for $55 each.

"Unlike non-AI malware, it could figure out unfamiliar devices on its own."

https://www.anthropic.com/research/self-replicating-ai