> Most (in fact all) pesticides used are designed to kill biological cells

You can make pesticides based on hormones of the target insect. Those pesticides do not kill cells.

Insects are kind of like little biological state machines, with hormones controlling the state transitions. A pesticide based on those hormones can mess up the timing. For instance, suppose you have in insect the munches on your crops all summer, then when it gets cooler and wetter lays its eggs and dies, leaving the eggs to repeat the cycle next year. A hormone-based pesticide might be able to make them lay the eggs early, when it is too warm and dry for the eggs to survive and when things that might eat the eggs are active.

I can't state how wrong you are. Believe it or not, different classes of life have different cells with different DNA. They use different pathways for almost everything.

As such, they are not equal. Specifically, insect cells are generally "more advanced" than ours (have more recent "basic" genes, more accurately adapted to their environment, tougher, ...), a lot smaller, ...

Bacteria are far more different from us than that, as are spores, other plants, ...

So yes, there are a LOT of compounds that throw a wrench into, say, insect procreation but have no known effect on humans.

> So pesticides do not distinguish between a caterpillars cell or a humans.

Maybe not all of them, but many, such as Bt-derived toxins, are famous for it.