Additive manufacturing has been losing a fair comparison for fifteen years: a part designed for milling is better manufactured on a milling machine. Flat faces for clamping, straight holes because the drill bit is straight, draft angles. That vocabulary is so ingrained that it's no longer visible.
An optimizer doesn't have that. You give it the available space, the anchors, and the loads, and it returns the material distribution that supports that with the minimum mass. What comes out resembles a bone: ribs that follow the load, gaps where material wasn't needed, sections that change continuously.
And that way can't be done any other way.
It cannot be molded because there is no draft direction. It cannot be machined because there is nowhere for the tool to enter. It cannot be cast without cores, which are themselves impossible. The part that is impossible for all other processes is exactly the one that an LPBF machine manufactures effortlessly.
What will multiply this is not the power of the optimizer—it's been in simulation software for over a decade—but the level of expertise required to use it. The day describing the problem is enough to obtain a reasonable design, we'll be optimizing parts that no one even considers optimizing today. Each one is an application of this technology that isn't currently on any list.
With its fine print, which we also discuss: optimization isn't always profitable, organic doesn't mean manufacturable, and a closed-cell lattice is a dusty part that will never be removed. The whole argument, with its four caveats.
