It's about the origin of life. Waaaay back in the middle of the 20th century, Urey & Miller did some famous experiments where they showed that you can take chemicals from the early Earth's atmosphere, zot them with some electricity, and get amino acids and other organic molecules that are the building blocks of life.
The general scientific picture that then developed was that inorganic processes like lightning and UV radiation and so on create this primordial soup of organic chemicals, and at some point, by random chance, some of the chemicals come together to form a self-replicating system, and then everything else is an outgrowth of that one pivotal random event. Everything about life, the whole shebang, is basically a "frozen accident". Later Nobel-winning research by folks like Tom Cech simplifies things a little by suggesting RNA as the original replicator, because it can be both a product and a catalyst, so now you have an RNA-based world for a while before DNA develops, but it's still basically the same setup.
The AmSci article proposes that things went a little differently. The authors call it "metabolism first" and it basically boils down to this: chemical reactions can experience evolutionary selection well before you get anything that looks like a self-contained replicator. You don't have to start with something complicated like RNA -- that's a much later elaboration of the basic replicator, which is just a sequence of chemical reactions. You don't even need cells; this network of interactions happens in the cracks of porous rocks, or maybe in a layer of organic goop on the surface. It's nothing more than a collection of small molecules that catalyze the breakdown of high-energy molecules into lower-energy ones, making more of the catalyst molecules as a byproduct.
The analogy they give is a pool of water up at the top of a hill. Water wants to run downhill. When the pool starts to overflow, at first, you just get a small trickle of water coming over the edge, but that flow digs the channel a little deeper, which lets more water flow, and pretty soon the erosive feedback has cut a deep channel. The basic driver of the flow is the energy imbalance of all that water at the top of a hill. The primordial environment had an energy imbalance as well: ordinary geochemical processes produce hydrogen and carbon dioxide, which have lots of electrons in a high-energy state. You can lower their energy by combining H2 and CO2 to form water and acetate, but the reaction is very slow. So the electrons in those molecules are basically stuck at the top of a thermodynamic hill. Add some catalyst molecules, and a few electrons can flow "downhill" to a lower-energy state. If you produce more catalyst molecules in the process, it carves the channel a little deeper, more electrons flow downhill, and feedback amplification has the whole show off and running. (The thing that makes it complicated is that there are actually a bunch of different channels, and flow through channel A doesn't dig itself deeper, it deepens channel B, which connects two other pools halfway down the hill. The important thing is that the system as a whole is self-amplifying.)
What did this primitive chemical network look like? Well, at the core of all metabolism is the citric acid cycle, which you may remember from high-school biology as the Krebs cycle. It's a simple circular loop of reactions that's the starting point for all the biomolecular synthesis pathways. In the usual "forward" or oxidative mode, it uses oxygen to break down organic molecules into CO2 and water, creating high-energy electrons that get ferried off to drive other biochemical reactions. However, it can also run in reverse, in reductive mode, taking in geologically-produced high-energy electrons and using that energy to create complex organic molecules directly. And that's exactly how some anaerobic organisms use it. Given its simplicity, universal centrality, and match to the early environment, the authors of the article think that the citric acid cycle is probably where it all started.
So that's all kind of interesting, but it's not what makes the new idea really cool. It's the philosophical implications.
The problem with the standard view of life as a "frozen accident" is that it relies on chance. It was just random luck that at some point, various molecules happened to come together in just the right way to produce something interesting that persisted and grew.
This is... dissatisfying from a scientific point of view. It's aesthetically discordant with what we have learned about how the world works. Science is all about reproducibility and universality, and recognizing that things here are the same as things there because the rules are universal. Special circumstances are not required to explain what we see. Now, that doesn't mean it's not true; given enough throws of the dice, even very unlikely events become probable, so in a large enough universe, that lucky accident was bound to happen somewhere, and wherever it was, that's where we'd be to see it, right? But still, the notion... itches.
Metabolism First turns that whole picture on its head. In this scenario, biology is not a fortuitous accident of chemistry, fighting against the thermodynamic odds to come into existence; rather, it is an utterly unsurprising, perhaps inevitable consequence of the geochemical circumstances of the primordial earth. Biology isn't a superfluous result, it's a necessary one.
From a thermodynamic perspective, life can look a bit unlikely. The arrow of time disorders things and smooths them out, while living systems are these complicated lumps of lowered entropy. But that's just near equilibrium; very far from equilibrium, the arrow points the other direction, creating localized order that speeds the flow downhill toward balance.
Under Metabolism First, life isn't weird at all. Life is totally normal. We should expect life, or something like it, to arise wherever we find circumstances like those of the early Earth. Anywhere you have wide-ranging conditions of strong metastability, thermodynamics will drive complexifying life-like processes to move it toward equilibrium.
We are not special or unusual; the Earth does not hold a unique position at the center of a barren and empty universe. Everything is like this. This is how things work. We're just one expression of a cosmic pattern encoded deep in the very laws of reality. The universe plays dice, but the dice are loaded. The system wants to be interesting.
Wouldn't that be a fascinating universe to live in? I think that it may well be where we do live, which is just... awesome. In every sense of the word.