What changes the outcome of the experiment, is the interactions that are given priority within the experiment…
Which become important in quantum physics, because quantum operators do not behave like classical operators. Allow me to provide an explanation that should not be beyond anyone’s pay grade here.
If you have two classical physical operators, say, a classical position and a classical momentum operator, then the order in which these operators act within a classical system is not important. The operators are said to commute. Let our operators be labelled x and p, say. In a classical system, applying x first then p (xp), yields the same result as applying p first then x (px)…
There’s a special mathematical construct, called the commutator, that is defined as:
[x, p] = xp -px
For classical operators that commute, this commutator construct is always zero.
Quantum operators, on the other hand, do not commute. The order of application of those operators to a physical system is important, and for the quantum versions of those operators, x and p, the commutator:
[x, p]
is non-zero.
Now while a scientist in a laboratory can, of course, affect the order in which those operators apply in an experiment, you don’t need a conscious entity to be present to affect the order. Any physically permitted process may have the same effect, even without intervention by a sentient entity.
It took time for some physicists to realise this, because, by definition, experiments are subject to human control, and devising a means of experimentally testing operator order arising without human intervention is a non-trivial problem in quantum physics.
I’m reminded here of an experiment to try and circumvent Heisenberg uncertainty using rubidium atoms. You can, if you wish, make rubidium atoms undergo interference (courtesy of the fact that they have a De Broglie wavelength associated with them - a very short one, I might add!). But, you can also target those rubidium atoms with a microwave laser, and excite them. Doing so provides a change in momentum, but far too small, in theory at least, to alter the path of those heavy atoms.
The reasoning was that you could identify which of the rubidium atoms reaching the screen after two slit interference were excited by the microwave laser, and thus use this information to determine, via a roundabout route, position and momentum not subject to Heisenberg uncertainty.
BUT … when this was tried out, it failed. Turn on the laser and excite rubidium atoms passing through one slit, and the interference pattern vanished. Only when the microwave laser was turned off, did the interference pattern resume.
That experiment led to the development of the concept of quantum entanglement, which, wait for it. is defined in terms of the operators having a non-zero commutator. 
Enjoy.