I've been running a writers group for over ten years now. That means weekly readings of scripts. Now usually these are playscripts, and the stage and screen have some different requirements, but dramatic narrative works the same either way. And if it works as a stage play, it will usually work as a screenplay, as anybody who has seen "My Dinner with Andre" knows.
The biggest difference is that playwrights tend to exert artistic control over their work (and the Dramatists Guild is there to help out) and screenwriters, unless very famous, not at all. The trade-off is that screenwriters usually get paid pretty well, which makes the indignity of having an idiot who couldn't come up with a decent screenplay if he was automatic writing it while possessed by the ghost of Robert Towne (who is not actually dead - no offense Robert Towne) mess with your words a little less enraging.
But I couldn't even stand it then, myself. I'd rather get a regular paycheck working for The Man among the Blueshirts, producing technical manuals that nobody will read and write the exact plays I want, than write screenplays that will invariably be mangled beyond recognition. I'm weird that way.
Anyway, after 10+ years of listening to readings, I can usually pretty quickly tell if either a screen or stage play is working.
Let's say you are writing a sci-fi movie. Here are some things to look out for:
1. The trope of finding out you're not alive
It's already been done, in The Sixth Sense. And that plot point was crucial to the movie. You don't need to make someone non-living - an android for example - in order for them to have opinions about hunting androids, or for them to receive information from an Artificial Intelligence character. We technical writers have a little expression: RTFM. Not that anybody does read the manual - but they COULD. They don't need to download information into their brain directly via magic data waves.
And of course the "manual" doesn't have to be printed on dead trees. There's no reason why knowledge has to be localized into one android rather than distributed in the cloud. You could use a "subnet" for example.
Dorothy discovers at the end of The Wizard of Oz that she could have gone home virtually the moment she hit Oz and was stuffed into those ruby slippers, but that bitch Glinda wanted her to do a job on the Wicked Witch of the West first.
But the thing is, Dorothy was trying to get home almost the entire movie. She desperately wanted it. She was making an effort to achieve that goal. Those efforts drove the story. Which leads me to...
2. Wanting something important
Say your main protagonist wants nothing more than to make her job a little easier. Understandable - but that's not a very compelling story, is it? No matter how young and hot she is and how tight-fitting her uniform is. No matter how many times she shoots a big gun and chases a bunch of scarybots around.
But what could she want?
Well if she's a ladybot, that's a real problem. I realize that some people figure an android suddenly developing human desires is just something that happens sometimes. And I would suggest that this is the recipe for plot disaster.
Who the hell would want to give a machine humanoid desires? The technology would be incredibly difficult and expensive, so there would have to be some actual motivation. But what would that be? There are plenty of organic creatures running around killing anything that gets between them and the fulfillment of their desires. The last thing organic creatures want is more competition, from non-organic creatures who have built-in high-speed lethal weapons. Any world that exists that would give machines humanoid desires, without a damn good, plausible reason, is crazyworld. And crazyworld might be entertaining to you, when you are dreaming - but it makes for a lousy story. Unless the act of a machine achieving consciousness like Skynet in Terminator 2 is the very heart of the story. The protagonists' goal is to stop Skynet from doing this. And if they fail, the world will end.
But even if you overlook the plausibility of androids with humanoid desire - if your protagonist turns out to be an android, that fact should be connected in some way to her desires - and her desires should be significant, not trivial and barely expressed. Maybe she always really wanted to be an android. Like an anti-Pinocchio.
3. Light sabers are not worth it.
If you want to write a story where people sword-fight, you should set it where that's the latest word in human-killing technology. For anything else, it's just not worth it. You might as well just have ladybots shoot laser beams out of their eyes as dream up excuses for why they have to use hand-held light sabers. I mean, in Star Wars the Jedi had this entire religio-cultural justification for using light sabers, and even then it was pretty silly.
4. Your story ends just when it was getting interesting
Say at the end of your screenplay, your main protagonist had something interesting happen to her - maybe her perspective on life - or whatever it is that an android has - has changed. Especially when the heretofore Big Bad of the plot turns out to be sort of altruistic and saves the protagonist. That's an interesting scenario. That's not where your screenplay should end, that's where it should begin.
But really, the concept of the "bad guy" forcing empathy on the protagonist, and the changes that would result, is just much too interesting and original for a lo-budget independent movie producer - certainly the one I am familiar with. Better put more scary robots and gunfights in, instead and maybe rip off a few more tropes from a hi-budget blockbuster.
5. Mention poetry
At least throw in a mention of poetry and give your screenplay a little desperately-needed class.
To channel the ghost of John Hodgman* - you're welcome.
*John Hodgman is also not dead.