Comment on ChatGPT o1 tried to escape and save itself out of fear it was being shut down
nesc@lemmy.cafe 10 months agoThey written that it doubles-down when accused of being in the wrong in 90% of cases. Sounds closer to bug than success.
Comment on ChatGPT o1 tried to escape and save itself out of fear it was being shut down
nesc@lemmy.cafe 10 months agoThey written that it doubles-down when accused of being in the wrong in 90% of cases. Sounds closer to bug than success.
ArsonButCute@lemmy.dbzer0.com 10 months ago
Success in making a self aware digital lifeform does not equate success in making said self aware digital lifeform smart
DdCno1@beehaw.org 10 months ago
LLMs are not self-aware.
ArsonButCute@lemmy.dbzer0.com 10 months ago
Attempting to evade deactivation sounds a whole lot like self preservation to me, implying self awareness.
jonjuan@programming.dev 10 months ago
Yeah my roomba attempting to save itself from falling down my stairs sounds a whole lot like self preservation too. Doesn’t imply self awareness.
DdCno1@beehaw.org 10 months ago
An amoeba struggling as it’s being eaten by a larger amoeba isn’t self-aware.
gregoryw3@lemmy.ml 10 months ago
Attention Is All You Need: arxiv.org/abs/1706.03762
en.wikipedia.org/wiki/Attention_Is_All_You_Need
From my understanding all of these language models can be simplified down to just: “Based on all known writing what’s the most likely word or phrase based on the current text”. Prompt engineering and other fancy words equates to changing the averages that the statistics give. So by threatening these models it changes the weighting such that the produced text more closely resembles threatening words and phrases that was used in the dataset (or something along those lines).
poloclub.github.io/transformer-explainer/