AI models may be capable of self-replicating without access to a datacenter

Posted by The_Lord_Humungus

5 Comments

  1. Another “safety experiment” that can be summarized as “we made AI do a bad thing and it did the bad thing”.

  2. MaximumEffort433 on

    >Our experiments suggest that small AI models may be months away from being capable enough to self-replicate on commercial internet-of-things (IoT) devices, like smart ovens and off-brand Android TV boxes. A replicator like this would be challenging to shut down, since there’s no centralized point to shut off—though it may be slow (with a doubling time of days). Unlike non-AI malware, it could also figure out unfamiliar devices on its own. In our evaluations, a 9B model got inference running on 8 of the 9 hardware types we tested.

    Oh good, that’s just fantastic. Adaptive malware that can spread itself to other devices, that’s exactly what I wanted for Christmas, constantly evolving digital kudzu.

    Remind me again why we aren’t shutting this shit down in its crib?

  3. RottenMilquetoast on

    Okay, hear me out: CyberPunk, but no cool fashion, no friendship with a rogues gallery of rebels, no hot cyborgs, no philosophically interesting sapient AI.

    Just drones, LLMs and twitter. 

Leave A Reply