Most educated people understand, at a basic level, what a cloud is and how it works. Water rises into the atmosphere and forms water droplets and ice pellets that appear to us as clouds. Water and ice eventually fall to the ground as rain and snow. Clouds dissipate. An average meteorologist could conjure a pretty convincing explanation for even the most awe-inspiring cloud formation without exceeding 200 words.

Don’t conflate a basic understanding with deep knowledge. The truth is that humans know very little about the goings on in the atmosphere, which is why predicting rain a couple days out remains a crap shoot. Our climate models basically work, but they leave out small yet important factors about atmospheric processes, like how crashing ocean waves and air pollution impact cloud formation.

That knowledge gap is sad because clouds are pretty powerful and very interesting. Many researchers have suggested clouds could be the key to reversing climate change. Just send a fleet of ships out into the ocean and spray salt water into the atmosphere, the thinking goes, and the increased cloud cover will send more of the sun’s energy reflecting back into space, leaving us a little cooler down below.

Still, there are good reasons why we haven’t actually tried this. Mainly there’s this: It’s really, really, hard to predict the unintended consequences of climate engineering at scale. It’s possible that the increased cloud cover would insulate the Earth better and actually result in a net heat gain for the planet. Obviously that would be a bad result.

And although we could simply stop our grand cloud-making scheme, there’s no guarantee that our actions wouldn’t have runaway impacts that are impossible to rein in once the process gets going. What would it do to global weather? Which parts of the world would be flooded, and which impacted by drought? Would storms intensify or chill out? That’s the thing about trying to control a chaotic system — it is by definition uncontrollable.

If you ask an atmospheric scientist about the greatest mysteries that the sky still holds, they will point to a dozen areas of research and say, “we need more data.” And it’s true that more information leads to better predictions and higher understanding.

But the nature of chaos is that it is unpredictable. Even if it were possible to build a perfect model, it would be impossible to say for sure what the weather would do tomorrow. After all, if you can’t predict if a butterfly will flap its wings in Brazil, you can’t say whether or not a tornado will be triggered in Texas.

The solution may be to leave the weather to the weather gods, and concentrate more on the things within our direct sphere of influence, like keeping more fossil fuels underground. Humans, after all, have a sordid past when it comes to our attempts to harness and control the forces of nature. That’s what got us into this pickle in the first place.