What does "efficient" mean in this context? Maybe plants don't produce anywhere near the same number of kgs of CO2 for each Joule of sunlight, but that doesn't really matter. Plants are self-replicating machines which produce carbon as a byproduct. What matters is capital investment, human effort, or hydrocarbon energy per kg of CO2 produced. I imagine that plants are far more efficient in that metric than any current artificial process.

"Efficient" here is amount of CO2 captured by area. Not energy input, and certainly not by capital invested, but possibly by labor too.

For energy conversion, both by area, energy input, labor, and quite possibly by capital invested too, but both are close on that last one, so there's no clear win.

Both carbon concentration and solar energy capture are improving really quickly, so that picture may change soon.

Solar energy has certainly gotten much, much cheaper compared to decades ago, but there's only so much cheaper left to go before it reaches market price of plain sheet glass.

Collection efficiency is occasionally climbing a percentage point or two, but that's not an open end like a gut feeling raised on Moore's law might suggest. Even the hypothetical 100% cell that will never exist would not be a single order of magnitude better than what we have.

> but there's only so much cheaper left to go before it reaches market price of plain sheet glass

Yeah, I wonder who will be the first to make them over nylon cloth. On retail, photovoltaics already passed glass in lots of places.

But it doesn't need a lot of change to make it unequivocally cheaper than plants. Just the little margin of what we can do with the current design suffices.