Open Source Risks
Seth and Jon discuss the risks associated with releasing model weights openly, highlighting how safety measures can be easily trained out, potentially leading to misuse by bad actors. The conversation delves into the complexities of open-source models, emphasizing the importance of considering societal impacts.In this clip
From this podcast
Last Week in AI
#138 - DALLE-3, YouAgent, Gemini, NExT-GPT, AI book labeling
Related Questions