Obviously (1) is possible, because for any particular configuration of atoms, it is theoretically possible for someone to assemble atoms together in that configuration. I could randomly put atoms together and, by sheer luck, assemble a living, breathing Abraham Lincoln. But since it happened by random chance, I wouldn't know for sure that this was really a living, breathing Abraham Lincoln. I might observe it for awhile and notice that, golly, it sure does seem to be a living, breathing Abraham Lincoln. But I could never rule out the possibility that it only looks like that initially and that it'll eventually collapse or fall into some infinite loop or something.
Intentionally building a machine and KNOWING that it has such-and-such properties of a thinking machine, is much more interesting. Such KNOWING is, itself, an act of thinking, and so there arises an interplay between the thinking of the created machine, and the thinking of the creator. Then you can start asking questions about whether the created machine, being a thinker, could also itself intentionally create another thinking machine, and so on.
If you're tired of the endless non-productive debate about whether thinking machines are possible, and you'd like to go to the next level, and take steps toward actually quantifying things, I'd invite you to read some papers I've written on the subject. To lessen the learning curve, here are some very approachable slides, and if you like those slides, you can click through to one of the papers. https://semitrivial.github.io/MeasuringIntelligence2019.pdf