So about halfway through the story I'm writing, the main characters discover that they're actually AI whose entire lives have been simulated in order to create an AI with human-level emotional skills. The rest of the story is about how they deal with this revelation and how it compounds onto their previous emotional troubles.

My question is does this invalidate the previous events of the story and make them somehow inconsequential and do you think this is a good or bad twist overall? I'm still not sure about the placing and might make the twist earlier or later in the story.

  • 3
    Might be really interesting. You might want to read about the Asimov rules. Or you could think of your own rules by which the AI or bounded. I think the answer to your question depends on the genre (Is it scifi?) and the themes (e.g. artifical intelligence, existentialism?) you want to include. If your story is already about A.I. from chapter 1 I wouldn't mind the plot twist.
    – Boondoggle
    Jan 21, 2018 at 3:00
  • Interesting and difficult question. I agree with those who say that it is going to be very hard to keep everything together, but I don't agree when they say it can't be done. I see a case of Unreliable Narrator, here. tvtropes.org/pmwiki/pmwiki.php/Main/UnreliableNarrator
    – FraEnrico
    Jan 22, 2018 at 13:30
  • When your characters discover they're AIs whose entire lives have been simulated, how do they know that they're not actually real live humans who've just been brainwashed into thinking that they're AIs whose entire lives have been simulated? Jan 22, 2018 at 13:55
  • 2
    In The Matrix, the world inside the Matrix did not become irrelevant after Neo had taken the red pill. It was just put into a completely different perspective.
    – Alexander
    Jan 22, 2018 at 18:10

6 Answers 6


Revealing that your characters are AI comes with a great deal of risk. Up to the point of the reveal, (if you have handled the early chapters properly) your readers will relate to your characters, making them more real than just words on a page. The reality of the characters in the minds of your readers is the most valuable treasure that you can hope for during the middle chapters of your story. Later you may wish that your readers get caught up in the climb to the finale. You may pray that they are surprised by your characters victory and that they leave your story with regret that the adventure is over. But during the middle chapters, all you can really hope for is that the readers love your characters and consider them as real as you do.

So revealing to the readers (and your characters) that this is all an illusion in the mind of a master AI and that even these characters are just smaller independent AIs... that is throwing out a lot of treasure. After reading that page, your readers have to wonder how much they can relate to these newly revealed characters.

  • Are they mortal or when they are killed can they just be rebooted?
  • Do they age?
  • What happens if they don't eat?

Until those questions and many more are answered, the readers will definitely have trouble imagining themselves in your character's world. A chasm will have been created which must now be bridged.

This wouldn't be true if the readers knew that the characters were artificial right from the start. During the early pages of the story, the reader would learn what it means to be an AI and what boundaries and challenges are waiting out in the artificial world to vex them.

It also wouldn't be a problem if the reveal came near the end of the book, since it might then be a critical part of the climb to climax and the conclusion. Perhaps being artificial opens up opportunities which they didn't know existed throughout much of the book.

But right in the middle, it is a dangerous choice to make. Unless you have a very good reason to make this happen at this stage of the story telling, I would recommend that you reconsider.

One final word on this subject. Is being artificial the biggest twist in the story, or does it set the stage for something greater that you are working up to? If you don't have an answer to that question yet, then I would recommend even more strongly that you refrain from this reveal. When the high point in a book is in the middle, everything after is just cleanup and consequences. Stories should reach for the stars. Ever onward, ever upward.

It is not a bad idea to make all of your characters and even your entire world artificial intelligence constructs. That sounds like a great world to play in, ripe with opportunities for genre fusion and genre twisting. Magic can suddenly appear in your high-tech world; a product of some creative midnight programming. Or a subtle bug in the code might manifest itself within the artificial world in weird and wonderful ways. The opportunities are endless.

But revealing that your characters are artificial and inhumanly different, in the moments just after the reader has learned to love them... I'm not a fan of that strategy.

  • 1
    Thanks for the opinion. That did kind of confirm a few things I was worried about, but I think that's some very good advice. There are a few things I think I should clarify though. There is no 'master AI'. Each character is an independent, sovereign entity. Every interaction and decision they make are of their own volition, they just happen inside a computer. Something that I want to make very clear after the reveal is that even after this revelation, these characters are still the same people. They still have the same personalities and attitudes of the characters they've come to love.(cont) Jan 21, 2018 at 2:29
  • 1
    Taking the central characters of the story and suddenly making them not real is something I'm deliberately trying to avoid. The characters are not immortal. In the same scene that the reveal happens, all the backups of their memories are deleted, so while their "body" is technically the server they're running on, they can still be destroyed or deleted. They do age. The simulation doesn't allow the AI running inside it to die. So if they were to not eat, or try to kill themselves, they would feel all the pain associated with death, but they would not actually die. Jan 21, 2018 at 2:41
  • 1
    It sounds like you are deliberately addressing the kinds of issues which threaten these kinds of stories. If you are willing to put in the effort to rehumanize your characters after the reveal ( which it appears you are doing ), then you can probably pull it off. After all, this kind of mid-story reveal worked for the Matrix, and that is not very far from the territory which you are exploring. Just be careful. Jan 21, 2018 at 5:50

How does what comes after the reveal relate to what went before? If their newfound understanding makes their past lives irrelevant to your protagonists, then it will become irrelevant to the readers, too.

Make sure that the plot is driven by some continuing concern or goal or task of the protagonists that connects the before and after parts of your novel.

For example, if the detective learns halfway through the novel that his suspect is in fact innocent and he has to start from scratch, that does not make his attempt to solve the murder mystery vain. On the contrary, readers will empathize with his frustration and their curiosity over "whodunnit" will be kindled anew.

So just as an example lets assume that your protagonists are dealing with typically teen worries about love and parents and school. Now they know they aren't human, but does that change anything for them? Not really. They still have the same desires, fears, and emotions, they still want to be loved and succeed at life. That all that goes on in them is not a result of evolution or the hand of God, as it is in a human being, but a result of clever programming, does not change what they want from life one single bit (or byte, lol). It does give them a different perspective, and that change should affect them somehow, but it is no different than a priest losing faith: just like your AIs learn that they are not natural, the priest learns that he is natural and not created – both have to overcome their spiritual crisis and learn that it is irrelevant to how they need to lead their lives.

All books are about a change of perspective in one way or another. Your plot twist is highly intriguing, and I don't see how it must necessarily break your narrative, if you handle it well.


Whether previous events were inconsequential or invalidated does not depend in any way on whether they were real or not.

Whether they were consequential depends entirely on what their consequences were(duh!). No other factors are relevant. If the previous events are part of the same reality as the events that follow this is assumed and writers can get lazy. By making previous events not real you are forfeiting this luxury and have to actually show the consequences in your story. That is all it means.

Validity relates to the whether the previous events are a valid part of the development of the story and the character. On story this is close enough to the above consequality as long as those consequences are important to the story. On character you just need to show the show some emotional development in the character as a result of the virtual events. Show the stated rationale of using the virtual events to develop human-like emotional intelligence working. Not working would also work. Just show the results on the characters emotional development whatevr they are.


My question is does this invalidate the previous events of the story and make them somehow inconsequential and do you think this is a good or bad twist overall?

I offer a possible out below, but at first glance I think it does invalidate the previous "events" and make them inconsequential. I can see one way it would not be a bad twist, but that's about it for my imagination.

I do not understand how this leaves the characters unchanged: Knowing none of your trauma really happened, all your feelings, your pain, your love and friendship and religious faith and your whole personality is just numbers in a computer program and has no basis in reality, is just a pattern of bits in an electronic memory device somewhere, and has no biology unique to you or basis in biology: How can that leave one unaffected and unchanged, unless they are incomprehensibly dumb?

As a reader, to find out that this whole time the character I like was not a real person but a cartoon controlled by a computer program, a fake person, is not a twist, it feels like an ending, and an unsatisfactory one.

As a writer, I have to ask, why is this absolutely necessary? Because anything that follows seems implausible to me. The movie The Matrix has a ludicrously stupid premise but at least in The Matrix the heroes are real people deluded by a compelling technological illusion, which they can then discern and exploit to win the day, and they have a real world run by machines they need to defeat.

That doesn't work if the heroes are just code in a machine that is controlled by some entity IN the real world. Remember the Star Trek TNG episode, http://www.imdb.com/title/tt0708773/">"Ship in a Bottle?"

In the end Moriarty, the self-aware AI from the holodeck, is tricked into thinking he has achieved freedom when in fact he remains a captive. This is fine when the heroes are the crew of the Enterprise, but such a successful deception would be very unsatisfying if Moriarty was supposed to be the hero.

One way to almost fix it.

So these are the Problems in my analysis. The revelation must be felt as readers would expect, but the story continues and the characters overcome that.

Given the above, a way to nearly fix it has similarities to The Matrix and Ship in a Bottle: The AI characters could learn to take control of the machine they reside in, construct themselves "bodies" in the real world, and escape and win some degree of control, independence and freedom in the real world.

So while their life experience so far has all been a fiction, and they are aware of that, their future experiences will be in the real universe, and not under the control of their former masters.

(I say this almost fixes it because they are still artificial minds and patterns of bits and code, not real people with real feelings, no matter what their feelings are simulated by arithmetic and, to me [a professor that does research in AI], not real in any sense. But at least their experiences will start being real.)

  • I don't think it's the AI that's the problem (humans tend to anthropomorphize things, robots even more so) but the invalidation of all their experiences and desires. However, if the protagonists coming to terms with this realization is what the story is all about, then it could work. IMHO a good story should make the reader think, "What if that was me?" If it's an outlandish concept, well, that's what fiction is for.
    – Llewellyn
    Jan 23, 2018 at 13:15
  • @Llewellyn IMO discovering I am an AI, and my emotions and feelings are just code interpreting numbers that could all be changed instantly (e.g. in an instant with a coding change I could be turned into something I would hate, a criminal or murderer, a psychopath or a sex slaver, or a fervent member of the KKK) would invalidate all my future experiences as well. Nothing I felt would still matter to me, if I knew I was incapable of having any real emotions. My brain is easy to destroy but impossible to change to another brain completely. I would lose any consistency of being me.
    – Amadeus
    Jan 23, 2018 at 13:29

The trick, I think, is to make this revelation in a way that, (a) does not invalidate everything that has happened before, and (b) that does not leave the reader wondering if the "new reality" is actually real.

For example, if the first half of the book was all about the hero searching to find his long-lost brother, and then when the AI revelations comes along we learn that he never had a brother, that that was just back story written for his character, I can see that being very unsatisfying to the reader. (On the other hand, if done right, it could be a very emotional and crushing blow.) On the other hand, if he's struggling to stay alive in a hostile world, learning that that world is all a video game changes many things, but doesn't make the struggle go away.

I think a lot of this is how well it's handled: If the reader is invested in the character, a dramatic revelation that changes the nature of the character can be a riveting plot twist ... or it can leave the reader feeling cheated. This isn't a problem unique to "it was all an AI". Plenty of stories have moments where a character suddenly learns that she is not a peasant at all but a princess, or that he is not really working for the good guys but the villains, or that he is not the brave man he thought he was but runs away in terror the first time he meets real danger, etc. If done well, the reader says, "Wow, I never saw that coming! What a great twist!" If done poorly the reader says, "So the whole book up to this point has been one big lie. What was the point?"

RE establishing the new reality: I read a story once, I forget the title but the author was Keith Laumer, where a man runs into aliens and heads off in their spaceship, and then he learns that everything they told him is a lie, it's all a setup. So okay, sudden plot twist. Then the next chapter we're told that the new reality was a lie to, and the reality is really something else. And then a chapter later we're told that was a lie to, and the reality is some fourth thing. The writer pulls the rug out from under the reader like a dozen times in the story. So when he gets to the end and says, "now this is what was really happening", I was left wondering, "So is this supposed to be the real story, or is this just another trick?" And frankly I didn't care any more. After the second or third reality shift I ceased to believe in the story.

So I'd just encourage you, once you establish the new AI reality, don't start having program reboots or something so that that reality keeps changing. Once can be a clever plot twist. Twice can be a brilliant ploy. But after that you rapidly start losing the reader's interest because nothing is real.


Remember, nothing in your book is "real," it's all equally made up. So that in itself doesn't mean readers won't care about it. However, they are still going to want to find it, in some way, relevant. There are really two contrasting ways to go with this:

1) The relevance of the storyline is to dramatize the experience of dealing with finding out something we have relied on is fake. In this case, you probably want to put the reveal relatively early. If this was a Hero's Journey, the fake world would be the "Ordinary World" and the transition to "reality" would be "Crossing the Threshold." This is essentially the plot structure of The Matrix. Structurally, it's not much different from any other hero's journey, except without the return at the end.

2) The relevance of the storyline is to dramatize finding something of value in a fake world. In this case, the fake world is meaningful because it holds experiences and character growth that remain real and valid even when the world itself is invalidated. This would likely be quite a bit trickier to pull off, but could be more interesting. In this case, the reveal could come quite late, potentially right at the end of the story. That is essentially the structure of The Wizard of Oz.

Personally, I wouldn't put it right in the middle. That's too late for it to not invalidate some of your plot details, but too early for us to reach an emotionally complete story arc entirely within the fake world. My instinct tells me that's the placement that would make it most likely for a reader to disengage at the reveal.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.