Blog/2026-07-23/A Dyson Sphere Misconception

ChatGPT's conception of what a Dyson sphere might look like. It would be dark in a cosmological sense but it has to radiate. And obviously the direction of observation counts.

It struck me the other day that I have a mental misconception of what a Dyson sphere would look like as it is built. To a sufficiently far enough observer, I imagined it as luminosity dropping until it 'winked out'. But you can't just constrain that energy. It needs to be radiated somewhere, and all the EM radiation coming from the enclosed star is just going into the shell of the object. So obviously what's going to happen is that over time the star would shift slowly into infrared as the shell is built until eventually it would be in the far infrared.

I'd recently recommended Linda Nagata's Inverted Frontier series, specifically the first book Edges, which got me thinking about the stars winking out as they become enclosed in Dyson spheres (not a spoiler, don't worry) and I realized that could not possibly happen. I tested this very original insight with an LLM only to find that the original article by Dyson actually says:

If the foregoing argument is accepted, then the search for extraterrestrial intelligent beings should not be confined to the neighborhood of visible stars. The most likely habitat for such beings would be a dark object, having a size comparable with the Earth's orbit, and a surface temperature of 200° to 300°K. Such a dark object would be radiating as copiously as the star which is hidden inside it, but the radiation would be in the far infrared, around 10 microns wavelength.

— Freeman Dyson, Science #131

Considering all that I say about citation laundering, I really should be reading more original material instead of simply relying on secondary references!