[PaulDavisThe1st]

Years ago I saw a nice comparison between optical computing and supersonic flight. The gist of it was that they would both only ever be widely used by the military/government. The specific benefits of both are real, but very costly and thus the use tends to be reserved where you absolutely need it.

For the general public, electronic computing and subsonic flight both work fine, are not tremendously expensive and don't require ongoing complex maintainance and engineering.

I would guess this analysis is still mostly true today.

[LyndsySimon]

To put it a slightly different way: existing technology is such that it is not profitable for them to be deployed at scale. I expect supersonic travel will eventually supplant our existing infrastructure (at least for long routes), but it's just not going to happen while it's multiple times more expensive to operate and has significant downsides like sonic booms over populated areas.

I believe the sonic boom problem has been at least partially solved, but the economics still don't work out. One day, they will. Likewise with optical computing - either ownership costs will fall to the level that they're acceptable for general use or advancements in parallel fields will make it impractical.

[selectodude]

Even with all the technology in the world, the laws of thermodynamics hold. Going faster raises friction losses requiring either a smaller plane or a larger amount of fuel. Either way, the price goes up. SST will never supplant subsonic transport on a large scale level, even accounting for the elimination of sonic booms.

[hn_throwaway_99]

> but the economics still don't work out. One day, they will.

This sounds like magical thinking to me. Supersonic travel will always be at a huge cost disadvantage to conventional airlines, and for the vast majority of people the benefits are not there.

I'm more curious about optical computing. I didn't realize there were costs in optical computing that make it fundamentally (i.e. at a physics level) more expensive than electronic computing; I thought the issue was mainly just that there was already over a half a century invested in electronic computing. What are the aspects of optical computing that make it inherently more expensive?