Think of it like this:

You can use a tennis ball machine to measure how far away a house is by firing the tennis ball at a constant velocity, timing how long it takes the tennis ball to come back to you, multiplying that time by the velocity, and dividing by 2 (since you measured the distance for a round trip). This works pretty darn well for measuring the distance to houses.

But now try this same trick to measure the distance to another ball. When your measuring ball hits the ball you want to measure, it doesn't stay resolutely planted in the ground like that nice friendly house. The energy from your measuring ball bounces the ball being measured off into the distance. Even if you could get your measuring ball to return, the ball you measured isn't in the place you measured it.

Replace that tennis ball with a photon, and you have the basic picture. There's no such thing as passive observation. Measuring something interacts with that thing. Conventional measurement is like in the case with the house, the thing being measured is so much bigger and more stable than the thing we're measuring with that the effect is negligible. But once you start trying to measure something on the same scale as your measuring tool, the ensuing chaos makes it basically impossible to get useful measurements.

Every road leads to Plato's cave

My teacher had a good comparison for this: observing macroscoping reality like we do microscopic reality would be like throwing a car at another car to measure its speed or position. Obviously you alter the course of events this way.

Fortunately light doesn't do much in the macroscopic world, so we can use it to observe stuff.

Great distinction.

What if you just measure the ambient particles