The atmosphere masses roughly 5.5x10^18 kg. Removing 140ppm to return to pre-industrial levels is 7.7x10^14kg of CO2 to split: 110 tonnes per human. 44g/mol and

The atmosphere masses roughly 5.5x10^18 kg. Removing 140ppm to return to pre-industrial levels is 7.7x10^14kg of CO2 to split: 110 tonnes per human. 44g/mol and a Gibbs free energy of -394kJ/mol, and we need 6.5x10^21 joules, or 1.8 billion GWh. More or less dead-on the energy in a trillion barrels of oil, or 10,000 1GW nuclear reactors working flat out for 20 years[1]. Something on the order of 2003 estimates of global petroleum reserves and also of gas reserves.[2] Sounds hopeless, but it's also a fraction of a percent of one year of total global insolation: about 12 hours' worth.[3] 1 trillion trees will also absorb roughly that much in about 30 years[4]. So not completely beyond the scope of physically possible human endeavour, but also probably substantially larger than any previous single endeavour. There's also nothing actually to stop the species doing something other than "it's expensive and hard and I get a better return from making an app for dog manicures". But money is a made up number and were actually not very short on human effort[5] or most materials or even usable energy[6]. [1]: in 2021, humanity operated 366 GW of installed nuclear capacity with 411 reactors. The line is currently going down (2020 had 369 GW and 415 reactors). The concrete and other materials presumably used in the reactors would mean there's a zero in the right half-plane: it'll get worse before it gets better, but nuclear is far from alone in this and by some estimates 50 times better than wind on a per-TWh basis. [2]: considering where the problematic carbon comes from, that's probably not a coincidental similarity in scale! [3]: and that shows in solar installed capacity: about 1TWh. Build about 10 times more than that, dedicate it all that new capacity to CO2 splitting, run it for 20 years, and party like it's 1499. It would need only 50 million acres: 10% of Saudi Arabia or Mexico would do, 1/3 of Texas or basically all of Great Britain (and a bit on the side to compensate for the latitude). Roughly 30m^2 per human: about $3000 worth of installed GW-scale solar power using an estimate of $2 billion/GW, plus the chemistry machinery and op-ex for 20 years. You might even negotiate a price break at the first trillion? [4]: but the wood has to stay in the tree or otherwise not be burned or rotted. Obviously avoiding burning down what trees we already have is probably a good start. [5]: other then then what will you do with all the "operationalised" humans needed to do it? Training that much of humanity would subvert a lot of existing political structures. If anything, this is why I think it won't happen under our current systems. [6]: and you'll get much if the material back afterwards, and what you definitely wouldn't be short of would be energy capacity!