I have a longstanding general interest in what might be term "unconventional computing architectures" (or at least "non mainstream computing architectures"). And, frankly, a general interest in "unconventional" things all around.
In regards to analog computers, it's that, plus an awareness of the constraints of the Von Neumann Bottleneck[1] and a vague notion that analog computing might (keyword, might) offer some potential to avoid that, and/or avoid the insane power consumption of modern digital computers (relative to a given workload).
And that in turn ties back to the Neuroscience book, as I'm reading that mainly because of my interest in neuromorphic computing[2]. And neuromorphic computing and analog computing, while separate topics, seem like they may have some interesting touch-points.
[1]: https://en.wikipedia.org/wiki/Von_Neumann_architecture#Von_N...