nprofile1qyt8wumn8ghj7un9d3shjtnyd968gmewwp6kytcqypzwj9j5zrphp5caje5l2qpa4gp26z45tljqqftahs47g42egf9l6l9sxvw (nprofile…sxvw) You're conflating warning / models. The post was a cautionary note, with Olma 3 as an example (I mentioned the superior performance of Claude).
Many of these are (OpenAI) positioned for social media slop / clicks / profits. Others (Musk: Grok) positioned for that + far-right extremist hate. They will be (with photorealistic video) 10-100x worse societally than the enormous disruptive impact of social media.