Conventional ways to do this include casting or, as is done on the RL-10 and SSME, building the nozzle out of thin-walled tubes (very labor-intensive and expensive). 3D printing is the future here, but there's still lots of improvement required in terms of wall thickness and roughness control, not to mention build volume.
Picture of blade and some process description: https://en.wikipedia.org/wiki/Turbine_blade
Picture of wax blade without cooling channels: https://en.wikipedia.org/wiki/Investment_casting
Company that uses DLP 3D printing for turbine blades: https://www.prodways.com/en/lost-wax-casting-of-turbine-blad...