@leroy is right in the sense that the overall trend of evolution is to less complexity. Track any moderately complex species over time and its descendant species are more likely to be less complex than the original.Not written like that, no. You're omitting a whole bunch of stuff and thereby misrepresenting it.
The last part "in some environments..." is also misleading, because it gives the impression that dropping any organism in such an environment would lead to a decrease of its complexity, which off course is simply not true.
What it actually means by saying that it favors simplicity over complexity, is more something like favoring efficiency.
For example, suppose some organism over time evolved complex defense mechanism X to deal with threat A.
As long as A is an active threat, natural selection will favor the presence of X, or improvement thereof.
Now imagine A goes away for whatever reason. Now, there no longer is a pressure to keep X working and intact. Instead, it will turn around and it will favor removing it entirely. It no longer gives the host an advantage and instead it just sucks away energy and resources for a system that no longer serves a survival purpose.
Another example is for example moles. They used to live above ground. Eyes are important there.
Now they live underground. Eyes no longer matter. Eyes have even become a hazard, as dirt can get in (especially underground) and cause infections, which might end up being lethal.
So the eyes are now hidden behind a thick layer of skin. Eyelids on steroids that are grown shut and can't be opened anymore. The eyes themselves also no longer work. So even cutting the skin open, wouldn't result in a mole with sight. One can imagine them eyeballs completely evolving away over time, as they no longer serve a purpose and instead just suck up energy and resources for nothing.
The point is, that evolution doesn't look for "complex solutions". Evolution rather looks of efficiency / efficient solutions. So at all times, it will always try and go for the simplest solution. The simplest solution, can still be a rise in complexity. There simply is no internal drive towards complexity. There is an internal drive towards simplicity. But there is an external drive towards complexity... The very nature of the problems, especially with life forms that already are simple, will almost always demand solutions that will end up in more complexity.
If you have NO defense against oxygen, acquiring such a defense will require a rise in complexity.
The need for defense mechanisms against an increasing variation of threats from competing populations, or environmental conditions, will also require a rise in complexity to evolve all those various defense mechanisms.
First life had NO defense mechanisms.
It is utterly unreasonable that life with simplicity on par with first life, would survive and thrive for 4 billion years.
Stephen Jay Gould gets into this a lot in his book Full House: the Spread of Excellence from Plato to Darwin.
If you plot out all species on a histogram of complexity, you also see from the skew that, in general, evolution favours less complex life.
Now... none of this suggests that complex life couldn't have evolved, just that the history of life is almost entirely about bacteria, and we humans are off in the tail of the distribution.