Dialogos reposted this
Self-learning AI alignment leads you down many rabbit holes: LessWrong, endless research papers, fringe blog posts, EA communities, niche subreddits, tpot, etc. When I first started exploring AI alignment in 2019, I was also falling into metaphysical inquiries that took a weird turn: pessimistic philosophy, depth psychoanalysis, the disillusionment of self. The two paths collided in ways I didn't expect, but I came out having learned that both technical and personal alignment require something coherent enough to care. I'm sharing my story of how I emerged out of that absurd hole along with Mettadology, a quarterly curation from Dialogos at the intersection of AI alignment, embodied cognition, and creative infrastructure. https://lnkd.in/ew6AqQDd