When ATP is used to power a biological process, it is converted to the diphosphate or monophosphate. Metabolic processes, such as glycolysis, then regenerate the triphosphate.
It's a bit like arguing that unloading containers from a cargo ship, and reloading it with new cargo, is the same as building a new ship.
"Made what work last night?
"The iterative machines!
"They worked before, what do you mean?
"No, not at this size. I re-designed the process and ran simulations until I had a new starting point. I replicated the first specimens and left them to themselves. They work now. They collect matter, they process it, they function! They REPLICATE! All it took was setting the penalty correctly.
"What scales of size are we talking about? What components did you use?
"Well, there are no components in that sense anymore. The level of integration is astounding! Everything that wasn't absolutely needed for function was removed by the optimizer and that what remained was functionally compressed and reuses as many existing 'components' as possible. That's what I meant. It required a severe penalty for the introduction of new 'components' and that is what did the trick. The results are self-selecting now. Specimen with more suitable parameters for the environment are able to out-replicate others and the succeeding ones were more likely to be based on successful, i.e. faster replicating ones. They improve themselves! And the results were orders of magnitude smaller than anticipated! This out-replication seems to have taken the role of selection. The things that work best will sometimes, by chance, create modified copies that work even better. Their faster self-replication makes them replace the old ensemble.
"That's prettty crazy.. How many orders of magnitude are we talking about?
"Haha.. see, the power source is a component that uses chemical compounds as fuel and extracts power from them. These are one of the larger parts of the machines and even then they are not even a micron across. A micron! And I wouldn't really call the rest of the parts components anymore. They are not clearly separated components anymore. The boundaries are fluid, a lot of parts are reused for different purposes, it is more one thing than a machine with different components. The whole things look like an awful mess! There is little discernable structure, heck, they don't even have a rigid structure anymore! The whole things can warp and move but they work. They work! They react to the environment! They can move in direction of a chemical gradient when they require the material for replication! We are talking about micro-machines. Heck, most of their parts are on a molecular level.
"Are you serious? That's madness. They are basically undetectable at these sizes. Have you any idea how dangerous this might be if they ever escaped your lab? How do you turn them off? What if they started converting available matter into copies of themselves?!
"Nah, they sure won't escape and if they do, the range of materials they can consume is too limited. It would be of concern if they managed - which I admit is a possibility because of their self-replication and the mixing of procedural parameters that occurs in the process - to increase the variety of input materials or worse, if they could harvest ambient energy in the form of heat or light.
"What do you mean by mixing of procedural parameters...?
"Yes, that was unfortunately required and makes the whole process somewhat unpredictable. Unfortunately - or maybe exactly fortunately? - I didn't get very far with intentionally designing these things. Once I went down the path of self-replication and introduced a way to not only self-replicate but letting them mix their own procedural parameters with the ones from other copies in the process, working results appeared really fast. I'm wondering if that might be the next level of machines? Us being no longer being assembled from components that were created in a plant and being loaded with a program but working machines from almost nothing! Starting out small, increasing themselves in mass and complexity until they reach a working condition. They are basically plants themselves.
"Spare me. How do you pause the process?
"Welll, that's sort of an issue so far. Since there are no separate components anymore, you can't really switch off the power. You can physically destroy them though or introduce substances that interfere with the machinery to disrupt the processes enough so that they permanently stop working. Or you withdraw the building materials but since they are such minuscule machines, they require very little and will work for quite a while. They deteriorate on their own with time if that's a relief. I do not know yet why this happens but after a while the processes slow down and most of the machines stop working shortly after that. On the other hand, they are frequently being replaced by succeeding generations which are usually more capable than earlier. It's definitely something we need to look out for.
Yes, it's totally off-topic but a short story I started a while ago. It is about a robo-engineer experimenting with microscopic machines and thereby creating what we call life.
> Us being no longer being assembled from components that were created in a plant and being loaded with a program but working machines from almost nothing!
This sentence needs some love, though. Had to read it a few times.
The format isn't uncommon, though. The rhythm of the first few lines is pretty strikingly similar and made me laugh a little.
But again, it's a relatively common format and rhythm. (And nicely written, I'll say)