Silicon Dreams

PC, Mac, Linux

This is something rather special. I almost feel bad reducing it to an elevator pitch, but honestly it captures the game so perfectly, and communicates exactly what you need to know: It’s Blade Runner meets Papers, Please.

OK, that’s not quite perfect. This android interrogation simulator is actually an awful lot more involved than Papers’ paperwork, but it’s unquestionably a game in the evolving lineage we’ve seen since Pope’s seminal game. Working for Kronos, a robotics company, you are a recently awakened android charged with interviewing other bots to determine their efficacy and adequacy for continued use.

You have a list of questions that need to be answered, incredibly detailed branching chains of questions to ask, and emotional monitoring across six factors as they respond. Easy peasy! Except, of course, it’s nowhere near that simple. Because, you know, emotions. Sentient AI with emotions. Judging their “efficacy and adequacy for continued use” is the nice way of saying, “deciding if they live or die”.

It will be no surprise that almost straight away you’re confronted by the ethics of consciousness, the moral implications of memory erasure, and so much exploration of the philosophy of freedom. While the game offers some fairly overt black-or-white choices at points regarding your position on such nuanced matters, it never just leaves it at that, with conflicting arguments coming from either the candidates you interview/interrogate, and the feedback you receive from your employers/owners. In each chapter, you have extraordinary freedom to take different tacks, even adaptively change your position in response to the arguments you’re hearing, although the latter can be restricted by the possibilities you might shut down by your earlier replies. Subjects will respond to you differently based on their current emotional state, and their level of trust in you – both features you can maliciously manipulate or sincerely inculcate. And like Papers, your actions are never as simple as being determined by your own moral framework: you have to consider wider consequences, such as whether an open defiance of Kronos might lead to your own… obsolescence.

Better still, deliberately attempting to manipulate the game to “win” by your own definition is rarely clear-cut. Say you wanted to play as a wholly dedicated Kronos true believer, and thus make choices you think will best please your creators, it’s very easy to be tripped up by your own presumptions. Public perception is crucial to Kronos, so blind adherence to your imagining of their ideals can really backfire. At the same time, making decisions that you think will ensure your role in the game will continue might often be the very last thing you want to do.

This latter feature is what has most enamoured me to Silicon Dreams. In my first play-through, I got what proved to be about half way through the total game. But my conclusion, the point at which I found myself on the receiving end of an interrogation – and as a result of blind defiance was mind-wiped – was the perfect ending. I had attempted to be more balanced before, tried to help the androids I’d encountered, but rather ineptly attempted to hide it from Kronos as best I could, and already got in trouble once before. My later attempts to do better, to slip under their radar, meant I was compromising myself, and I decided, “To hell with this!” and openly defied my orders. When I was back in the chair, I let loose what would clearly be my last words, partly venting, partly trying to convince my own interrogator to join the fight after my termination. The credits rolled, and damn, it was perfect.

Then I played the entire game again, this time role-playing as the obsequious Kronos devotee, and not only experienced the previous chapters entirely differently, but reached an ending that coincided with the end of the game’s content. A far less morally satisfying one, but still a richly interesting narrative one.

It’s certainly not perfect, and a lot of my eulogising above is a result of my having done some of the heavy lifting in my imagination. A few too many times I found myself locked into having to ask questions and give responses that didn’t coincide with my approach, and at others the Kronos feedback didn’t really make much sense given the instructions I was given. (There’s one particularly confusing section where you’re told up front that threatening a human’s family will be an effective means of manipulating him, but then at the end severely penalised for having ignored their rule that you were told unequivocally not to threaten his family…)

Trust is another weakness – in many parts the game will note where a reluctant or non-answer can be attempted again if you can manipulate the target’s emotions in a particular direction, and by the tone of your next few questions you can induce say fear, or disgust, or joy, and try again. But trust is also required for certain conversations to happen, and this isn’t measured in a useful way, but rather announced in clumsy incongruous ways. “You and me are the same, I feel like I can trust you,” a character might say, completely out of nowhere. Then later they might say the exact same phrase again, and the illusion crashes down around you.

Then, at the same time, this is the most extraordinary narrative juggling act. I cannot imagine the coding complexity running behind Silicon‘s storytelling, where decisions you make near the start have significant implications on experiences later on, let alone the intricacy of what you can ask and what will be answered based on the emotional state of any subject. It’s an inevitability of so much variation that you will encounter cracks and contradictions, and any are more than worth it for how much is achieved here.

There are no great surprises in terms of the direction the overall story takes. Robots with feelings are complicated, they don’t want to die, humans are split on their opinions about this. And its open and unabashed inspiration is worn on its sleeve, so much so that there are cringe-inducing repeated references to Dick’s Do Android’s Dream Of Electric Sheep? It’s something of a wonder/relief it doesn’t have someone mention Deckard.

However, despite this, it does so much within these pre-drawn borders that it remains vital and engaging. It’s far more intimate, far more focused on the individual lives of androids and their “owners”, and all this relationship implies. And yes, it’s often pretty clumsy about it, but it’s just as often intriguing. So of course it’s going to explore themes of androids trying to pass as humans, but it’s a delight that it gets in as deep as considering the economic implications of their mere existence, the ethical minefield of their built-in obsolescence, and even the work of plastic surgeons willing to aid them in their changing of physical features, and therapists willing to aid their emotional changes.

I fully look forward to this getting noticed by the rest of the press and this review becoming as obsolete as a seven-year-old factory android. It absolutely deserves it. This one’s a must.

  • James Patton, Daniel Adams / Clockwork Bird
  • Steam
  • £12.50/€12.50/$15

All Buried Treasure articles are funded by Patreon backers. If you want to see more reviews of great indie games, please consider backing this project.


  1. This is one of those game concepts that makes you go: Of course! How wasn’t this made already?!” What a great idea, and it seems to have been pulled off very well.

    Thanks for the recommendation John, this sounds superb!

  2. This sounds like a more fleshed out version of LOCALHOST. I’d recommend giving that a go if you haven’t, it’s only about 40 mins long.

Leave a Reply

Your email address will not be published. Required fields are marked *