The issue with considering these to be anything like the 'hard sciences' is that it is impossible to even try to control for all variables. Plus, whenever sociologists, for example, make a bad prediction, they just write it off as differences in personality or some other similar thing.

God forbid they actually just falsified their hypothesis. It's important that people understand how to think about the social sciences, don't get me wrong, but they're pretty overwhelmingly ineffective for creating a proper framework for understanding the world around you.

Theories in social science and theories in hard science are totally different.

Theories in science have a shit ton of evidence behind them and haven't been falsified.

Theories in social science, on the other hand, are all in competition with each other because they all have their positive and negative aspects that make them better for application in some situations than others.

And yes I know that we still use a newtonian idea of gravity in many cases, but that's completely different as it just tends to make the math easier in practice. It's not that we actually still believe in newtonian ideas.

Isn't one of the point of all those telescopes we built in space and on earth to prove or disprove our hypothesis regarding astronomy? Is that not experimentation?

No, the machines tend to be abstract. Such as an infinite paper tape that can manipulate symbols.

No, computer engineering tends to focus more on hardware. When I was doing that kind of thing in college, computer engineering did things like chip design and logic boards and so on. I had courses on DSP and VLSI, multiple assembly languages, RISC vs CISC systems, and so on. In my university, it was considered a subspecializqtion of electrical engineering, with the first two years of undergraduate study being identical.

When I switched over to CS, I was doing things like numerical analysis and software systems architecture.

Both majors used math, but CE (as an EE major) required students to go through (iirc) calculus 5, and I think that CS majors could stop at calc 3 but would end up having to do different kinds of math after that.

No, that's machines with math