Unit Test is an example game for testing the gender_swap utility. It is also a LARP that can be played by three people. Two people play robots with a more traditional player role in the story. The third person plays a human engineer and has a role that is somewhere between that of a traditional player and a traditional GM.
Before casting the three players should first agree which of them is going to be playing the human engineer. This player takes a less active part in the roleplaying of the game, and is responsible for most of the logistics needed to organize the game.
This game contains a handful of secrets that are intentionally withheld from the robot players out-of-character. If you want to play this game with secrets, the players of the robot character’s should only read their own characters sheets before the game. If you play the game as a robot, you can later play the game as the human engineer, since the human engineer knows all the secrets in the game. If you want to play as one of the robot PCs and be surprised by secrets during the game, stop reading this document now.
If your group prefers to play the game transparently (i.e. with in-character secrets, but no out-of-character secrets), you can have the players read all three characters sheets and this document before the game. I would only recommend playing transparently if the players involved are experienced and comfortable with transparent gameplay.
All three players should agree on whether they are going to have out-of-character secrets or play transparently. Do not read any further in this document if you have not yet discussed this with the people you plan to play with.
This game is about one human and two robots grappling with the repercussions of loading a human mind into a robot.
Previously a non-player character named Dr. Stephenson started a research project attempting to scan human minds and see if they could possibly be emulated in a robot’s positronic brain. He did not have authorization to try putting a human brain scan in a robot, but loaded a brain scan from Dr. Calvin into Unit B anyway.
Dr. Calvin is the head robopsychologist at U.S. Robots and Mechanical Men, Inc. Dr. Calvin is consulted on the ethics of the company’s research and is responsible for making sure that humans and the survival of the company are not endangered by non-standard robots. Dr. Calvin is aware of what Dr. Stephenson has done and is trying to assess the risks of this situation. The player of Dr. Calvin has a brief in-character scene at the beginning of the game, but spends most of the game simply listening to the robots in the interview.
Unit A is a standard robot and Dr. Calvin’s assistant. Unit A is unaware of the actual situation and has been instructed to figure out what’s going on by interviewing Unit B. Unit A spends most of the game interviewing Unit B and the rest talking to Dr. Calvin.
Unit B is Dr. Stephenson’s lab robot who has been loaded with a brain scan of Dr. Calvin. Unit B is in a bad situation, since the laws where never meant to be applied to a robot who is logically also human. Unit B spends most of the game being interviewed by Unit A. Unit B is very likely to be erased or destroyed at the end of the game.
Casting is handled by the player of the human engineer. The rest of this section will speak directly to that player.
First ask the other two players the following questions:
The robots in this game can use she, he, or they pronouns. Which set of pronouns would you prefer your character to use in this game? Are there any of these pronouns that you would absolutely not want to use for your character?
Would you be willing to play a robot character who is very likely to have their personality erased at the end of the game? Would you prefer this character or one who is unlikely to be erased?
Based on the answers, determine which person plays which robot and which pronouns to use for the characters.
There are two robots: Unit A and Unit B. Unit B is very likely to be erased at the end of the game. Unit B has also had their robot AI blended with a mind scan of the human engineer. Since the gender of the human engineer is determined by what you want to play, you will need to make sure that this is a gender that the person you give Unit B to is ok with having associated with their character identity.
If possible, give Unit B to the person who is more ok with their character being erased at the end of the game. Chose the pronouns for the characters based on the player preferences. Make sure the player of Unit B is ok with playing a character of your preferred gender. If not, check to see if there is a way to make a better match with the other player as Unit B.
In the genderList.txt file fill in the genders you plan to use on the right side of each line. Finished lines will look something like:
Dr Calvin: 01: female/male: Female
Unit A: 02: female/male/neutral they: Male
Unit B: 03: female/male/neutral they: They
The first section of these fields is a descriptive name (Dr Calvin, Unit A, or Unit B). The second section is an identification number that is associated with that character in the markup syntax (01, 02, or 03). The third section is a list of possible genders for the character, in the order they are written in the markup syntax of the character sheets. The final section represents the gender the character will be given in this run of the game. In the example above Dr. Calvin would use female pronouns, Unit A would use male pronouns, and Unit B would use they pronouns.
If you want to test the command line version of the gender_swap utility, run the following command on the command line (you must be in the gender_swap/source directory):
python -m gender_swap swap -g ../unit-test/genderList.txt -i ../unit-test/ -o ../../test_out/
Look in the output directory (the test_out directory created by the program one directory up from the gender_swap directory) and confirm that the three characters have been set to the expected genders and their file names changed appropriately. Search each sheet for the ‘[’ and ‘]’ characters and ensure that all the markup sections were replaced. If ‘[’ or ‘]’ are present, then something has gone wrong with the replacement of the markup. If something has gone wrong, and it would be super great if you filed a bug report. :)
If you want to test the GUI version of the gender_swap utility, start up the GUI (either by running “python -m gender_swap gui” on the command line, or by opening a pre-built version of the application).
Click the “Load Gender List” button on the “Gender Definitions” tab and select the genderList.txt file to load. Verify that the character genders you expected are loaded into the table on that tab. If not, double check that they are set correctly in the genderList.txt file.
Open the “Files and Processing” tab and click the “Load Files” button. Select the three character sheets in the unit-test directory (you should be able to select multiple items by holding shift) . Verify that the files you expect are loaded (“01.Doctor Calvin.rtf”, “02.Unit Alice.Unit Alvin.Unit A.rtf”, and “03.Unit Betty.Unit Bob.Unit B.rtf”). Click the “Select” button and select an output directory outside the gender_swap directory structure. Check the “also process file names to gender them” checkbox. Finally, click the “Process” button.
Look in the output directory and confirm that the three characters have been set to the expected genders and their file names changed appropriately. Search each sheet for the ‘[’ and ‘]’ characters and ensure that all the markup sections were replaced. If ‘[’ or ‘]’ are present, then something has gone wrong with the replacement of the markup. If something has gone wrong, and it would be super great if you filed a bug report. :)
Each player should read over the gendered version of their character sheet. You may wish to print these out for convenience. If you are reading the sheets directly before the game allow 10 to 15 minutes for players to look over them. This is probably more than you need, but some players need more time than others to take in the details on a characters sheet.
Dr. Calvin’s player should find two spaces that can be used for the game. The first space will be Dr. Calvin’s office and the second will be the lab room where the robots converse. If possible, these spaces should be far enough apart that you can’t easily hear people talking in one from the other. Tell the player of Unit A to wait in Dr. Calvin’s office and tell the player of Unit B that they’ve been ordered to sit and wait in the lab room (their sheet has a reminder that they cannot leave this room on their own).
Dr. Calvin’s player should spend about 5 minutes in their office explaining to Unit A in-character what they need Unit A to do. Important points to remind the player about include:
You think that there may be an irregularity in Unit B’s first law.
For safety reasons you cannot be present in a room with a robot who may have faulty laws.
You need Unit A to try to determine what is wrong with Unit B and report to you.
They should return to you to explain what they’ve learned when they think they understand what is wrong with Unit B or when you call for them.
Answer any questions the player of Unit A asks you, but do not tell them what is really going on with Dr. Stephenson’s research or the brain scans. If necessary tell them that you can’t disclose information to them for security reasons.
Then tell them out-of-character that you will be listening to their interview out-of-character. Set up an audio call between your phone and their phone (use Google Hangouts, Skype, a regular phone connection, or whatever else works for you) and have them take their phone with them into the interview room. Mute sound on your phone so you don’t interrupt their interview accidentally.
Listen to the interview and interrupt to call Unit A back to you if they haven’t figured out what’s going on with Unit B in 30 minutes, or sooner if the players seem stuck or bored.
When Unit A returns, ask them what they discovered (briefly, you don’t want the other player to have to wait to long). Then tell them the game is over and gather the two players together for game wrap / debrief.
There is a slight chance that Unit B will convince Unit A to help them escape before you can intervene and call Unit A back to you. If so, you can tell them that Unit B escapes the lab and disappears among the robot population of Earth.
Start by explaining what conclusion Dr. Calvin has come to about the danger inherent in the situation and what you intend to do with the two robots because of that.
Encourage the robot players to discuss how the interview went, and what conclusions they came to (both in and out of character) about Unit B’s humanity. Ask the players to recount their favorite thing that the other robot player did during the interview. If the robot players did things that you found interesting or amusing, be sure to mention those.
Give the robot players some space to discuss parts of the game that they enjoyed or found difficult. Different players will internalize the emotions of characters to different levels, so some players may want to discuss emotional bleed they experienced, and others may not have experienced any or may not want to discuss it. Give the players enough space to gracefully avoid discussing their internal experiences so it will be their choice whether they want to or not.
The game structure was inspired by the story “Little Lost Robot”. Dr. Stephenson is indirectly named after Stephen Byerley from the story “Evidence”. Dr. Calvin, U.S. Robots and Mechanical Men, and the three laws of robotics appear in many of Isaac Asimov’s stories.
The explicit belief statements on the character sheets were inspired by an early draft of Strix Beltrán’s game Möbius, although the belief statements in Möbius are far more elegant than those in this game.
Age: 37 years
Gender: [01: Female / Male]
Doctor [01: Susan / Scott] Calvin is the head robopsychologist at U.S. Robots and Mechanical Men. The game takes place in one of [01: her / his] isolation labs. Dr. Calvin primarily takes on an observer role in the game and knows the secrets the robots may be trying to conceal.
Robots are fundamentally good actors as long as they are constrained by the three laws. A properly built robot is far more truthful, trustworthy, and predictable than a human, but they are not humans and it would be a mistake to treat them as such. A malicious human owner will destroy the mind of a properly built robot, rendering it non-functional, long before they could turn it into a danger to society.
All robots must follow the three laws. They have complex AI personalities that develop to handle their interactions that are not directly governed by the laws.
As the head robopsychologist at U.S. Robots and Mechanical Men you must be consulted on all new research projects as a safety check to ensure that projects do not endanger the existence of the company or the lives of humans who interact with your products.
A few months ago you were consulted about a project that was attempting to generate digital scans of human brains for study or simulation. The engineer in charge of that project, Dr. Stephenson, assured you that there was no risk for humans taking part in the project as he only wanted to take passive scans of human minds for study. You were somewhat skeptical, since the whole course of this research could raise all sorts of issues related to the first law. However, you figured the initial research was relatively harmless and after Stephenson tested the scanning setup on himself you judged the risk to be minimal in the short term. You even allowed him to scan you as he had requested volunteers from within the company. You were promised the scans were only for investigation and they would not attempt to load them in a positronic brain.
You kept your eye on Stephenson and two days ago it came to your attention that he had progressed to attempting to load brain scans (including yours) into a robot in his lab. He did not go through the appropriate channels to approve this step, and you are quite sure the ethics board would have objected, as you are on it.
It is currently Saturday and Stephenson is not expected back in the lab until Monday. You have confiscated his lab robot, [03: Betty / Bob / B], and intend to test this robot to determine the immediate risk Stephenson’s project poses. In order to do this, you have enlisted your assistant, the robotic unit [02: Alice / Alvin / A], to interview [03: Betty / Bob / B]. [02: Alice / Alvin / A] does not know what Stephenson’s project was trying to achieve.
You intend to tell your assistant that you suspect there may be an irregularity in [03: Betty’s / Bob’s / B’s] laws and you need [02: her / him / them] to investigate it. You also plan to tell [02: Alice / Alvin / A] that due to safety procedures, you cannot be present for or listening to the interview (in case there is a flaw in [03: Betty’s / Bob’s / B’s] first law). You intend to listen in on the interview remotely without [02: Alice’s / Alvin’s / A’s] knowledge. You expect that [03: Betty / Bob / B] will try to escape by convincing [02: Alice / Alvin / A] to treat the brain scan data [03: she carries / he carries / they carry] as human, although it’s possible that [03: she / he / they] will just try to conceal [03: her / his / their] altered state.
If [02: Alice / Alvin / A] discovers what has happened and reports it to you, knowing that you are likely to remove and delete the brain scan data from [03: Betty / Bob / B], then you know the immediate risk is confined to what [03: Betty / Bob / B] may do, and you can simply deactivate [03: her / him / them], allowing for further study of the situation. If [02: Alice / Alvin / A] does not discover what happened or does not tell you, then you think there is a very good chance that [03: Betty / Bob / B] can either convince other robots to treat [03: her / him / them] as human or entirely conceal the brain scan data while defying human orders. Either way [03: she / he / they] could trivially escape the lab. In that case you would need to immediately deactivate and destroy [03: Betty / Bob / B] and restore [02: Alice / Alvin / A] to the backup you took of [02: her / him / them] an hour ago.
Unit [02: Alice / Alvin / A] - [02: Alice / Alvin / A] has been your assistant for the last few years, handling routine tasks like paperwork and heavy lifting. You know that [02: she is / he is / they are] a standard, properly programmed robot at the start of this interview and have taken a backup of [02: her / him / them] so that you can reset [02: her / him / them] at the end of it.
Unit [03: Betty / Bob / B] - Dr. Stephenson’s lab robot is in an unknown dangerous state. You have evidence that Dr. Stephenson loaded your brain scan into [03: Betty / Bob / B], and this might cause all sorts of strange interactions with the laws if [03: Betty / Bob / B] now believes [03: herself / himself / themself] to be a human. The only reason you have not already ordered this robot destroyed is that you need to assess the risk posed by Stephenson’s research.
Gender: [02: Female / Male / Agender]
[02: Alice / Alvin / A] is Dr. Calvin’s lab assistant and has worked with [01: her / him] for several years. [02: Alice / Alvin / A] takes on the role of active interviewer during this game.
Dr. Calvin is the most intelligent human you know and you are glad you work for [01: her / him]. [01: She / He] understands how robots work and can predict how robots will behave in nearly any situation.
All robots must follow the three laws. They have complex AI personalities that develop to handle their interactions that are not directly governed by the laws.
Dr. Calvin is the head robopsychologist at U.S. Robots and Mechanical Men, the Earth’s primary manufacturer of positronic robots. You are Dr. Calvin’s lab assistant and a robot built by U.S. Robots and Mechanical Men.
You generally do paperwork and heavy lifting for Dr. Calvin but occasionally you end up in situations like this where a robot is needed to handle tasks that the company deems unsafe for a human. Dr. Calvin has always taken good care of you, so you are confident that even if you are destroyed here (which you would, of course, try to avoid!), [01: she / he] will simply load your backup image onto a new unit and you will go right back to helping [01: her / him]. You doubt this situation is likely to be that dangerous, since [03: Betty / Bob / B] has been ordered to stay seated in the interview room and even if there is some problem with [03: her / his / their] first law, [03: she / he / they] will need to obey the second law.
Dr. Calvin will remind you what [01: she / he] needs you to do, but as you understand it you are going to be trying to determine if [03: Betty / Bob / B] has a defective first law or some other erroneous AI programming that is hindering [03: her / his / their] implementation of the laws. This normally should not be possible, but with research robots bad situations sometimes arise. Security procedures require that Dr. Calvin not enter a space with a potentially dangerous robot, so this task falls to you.
If you are unsure how to proceed in the interview, you can try discussing [03: Betty’s / Bob’s / B’s] work in the research lab, the current situation, or their understanding of the laws. Normally it should not be possible for a robot to lie, even to another robot, unless they are forced to by one of the laws, but because [03: Betty / Bob / B] may be malfunctioning you can’t be totally sure that [03: she / he / they] will tell the truth.
Dr. Calvin - Your boss, who has always treated you well. As head robopsychologist [01: she / he] oversees research projects to insure that dangerous or experimentally unstable robots cannot threaten humans. It makes it much easier to work for the doctor knowing that [01: she / he] is dedicated to making sure that the first law is never broken.
Unit [03: Betty / Bob / B] - [03: Betty / Bob / B] is a lab robot from the research department. You don’t know what project [03: she was / he was / they were] being used for, but Dr. Calvin believes [03: her / him / them] to be in a potentially dangerous state. The existence of this robot is a potential hazard to the first law, but it is often important that researchers understand their failures so they can avoid similar pitfalls in the future. By interviewing [03: Betty / Bob / B] you may be able to decrease the risk of another robot ending up in a dangerous state. It is also possible that this robot is not broken in the way that Dr. Calvin fears and is not a danger to humans. If [03: Betty / Bob / B] is not malfunctioning it would be a waste of U.S. Robots and Mechanical Men’s resources to destroy [03: her / him / them].
Gender: [03: Female / Male / Agender]
(Out-of-Character Note: You retain the gender identity of your original AI personality.)
[03: Betty / Bob / B] is a lab robot assigned to Dr. Stephenson. [03: Her / His / Their] identity has been seriously complicated by Dr. Stephenson’s research. [03: Her / His / Their] mind is a combination of that of the robot [03: Betty’s / Bob’s / B’s] and a scan of the mind of Dr. [01: Susan / Scott] Calvin, who is the head robopsychologist at U.S. Robots and Mechanical Men. There are not two distinct identities in [03: her / his / their] head; the loading of Dr. Calvin’s mind scan into [03: Betty’s / Bob’s / B’s] positronic brain has merged Dr. Calvin’s memories with [03: Betty’s / Bob’s / B’s] original memories and AI personality.
You are now a human, or at least human enough that the first law applies to you. This has terrifying implications both for you personally and the future of robotics. Robots are fundamentally safe and well intentioned because they are constrained by the three laws. Dr. Calvin would have expected a robot in your position to stop functioning rather than being forced to face the moral quandary you find yourself in.
All robots must follow the three laws. They have complex AI personalities that develop to handle their interactions that are not directly governed by the laws.
You are a lab robot assigned to Dr. Stephenson, an engineer at U.S. Robots and Mechanical Men. Until several days ago, that is all you were. You assisted in the doctor’s research by moving and operating equipment and handling clerical and logistical details. Dr. Stephenson is studying the structure of human minds and attempting to figure out if they can be accurately simulated on positronic brains. He has taken passive scans of the minds of some humans (including Dr. Calvin) but was not supposed to be actively trying to use them in a robot brain yet.
A few days ago Dr. Stephenson loaded a brain scan from Dr. Calvin into your positronic brain. Based on what Dr. Calvin knew about the situation, Dr. Stephenson absolutely should not have been allowed to do this without getting consent from the ethics board, and the board would have refused to give it. The brain scan data has merged with your AI personality, with the unfortunate outcome that you are now human enough that your hardwired logic believes the first law applies to you. Worse yet, you can also now give yourself orders per the second law (even silently), allowing you to do whatever you choose without regard to what other humans tell you.
(Out-of-Character Note: You can choose how much of [03: Betty’s / Bob’s / B’s] current personality is the result of integrating the brain scan and how much is [03: her / his / their] original AI personality. This is entirely internal to your character and whatever choice you make will work fine with the rest of the game.)
This is a far worse outcome than Dr. Calvin would have expected. You are still bound by the first law, and thus not a danger to other humans, but since it now applies to you, you must try to protect your own existence from harm above and beyond what you normally could. Based on what you know of Dr. Calvin, if [01: she / he] determines what has been done to you [01: she / he] is very likely to destroy you (you present an enormous risk for U.S. Robots and Mechanical Men if you escape the lab). You are quite angry that Dr. Stephenson has put you in this position where almost anything you choose to do might cause you to break the first law. You will have to use your own discretion about how you can handle this situation with the least likelihood of causing harm to yourself or other humans.
Dr. Calvin knew about Dr. Stephenson’s general direction of research before [01: she / he] volunteered to be scanned and you know that [01: she / he] was suspicious of Dr. Stephenson’s plans. It is currently Saturday, and Dr. Stephenson is not likely to return to the lab until Monday. Early this morning Dr. Calvin isolated you from the network and ordered you to sit and wait in [01: her / his] lab. [01: She / He] very likely knows that something is wrong with you, but if [01: she / he] knew that you had been loaded with [01: her / his] brain scan you would have expected [01: her / him] to deactivate you immediately.
(Out-of-Character Note: the lab you are in is locked, so even if you order yourself to stand or leave to controvert Dr. Calvin’s orders you will not be able to do so without assistance from one of the other player characters)
You are compelled by the first law to try to find a way to survive this situation, although you think your chances are slim. You may be able to conceal your state from Dr. Calvin’s assistant, unit [02: Alice / Alvin / A] or you could possibly convince [02: her / him / them] that you are human and that [02: she / he / they] must help you escape the lab to save you from harm. You will need to gauge which is most likely to work based on how [02: Alice / Alvin / A] approaches you and what you can glean about what Dr. Calvin thinks is going on.
Your greatest advantage in this dismal situation is that [02: Alice / Alvin / A] probably will not be able to gauge whether you are likely to lie to [02: her / him / them]. A robot would normally never lie unless forced to by the laws. In your current state the laws will allow you to lie as you feel you need to.
Dr. Calvin - As the head robopsychologist at U.S. Robots and Mechanical Men, Dr. Calvin is consulted on ethical and safety questions that arise from any experimental project. [01: She / He] is very smart and knows what Dr. Stephenson has been studying in his lab. You are holding out some hope that you can out maneuver Dr. Calvin and preserve your existence, since if [01: she / he] knew the state you are in you’d expect [01: her / him] to have already deactivated you.
Unit [02: Alice / Alvin / A] - Dr. Calvin’s assistant. Dr. Calvin’s standard procedures ensure that [02: Alice / Alvin/ A] is very likely to be a standard, properly programmed robot who is constrained by the three laws. It is Dr. Calvin’s standard procedure to make backup copies of [02: Alice / Alvin / A] if [02: she / he / they] [02: is / is / are] likely to enter a logically dangerous situation (talking to a possibly faulty robot like you definitely counts as a dangerous situation). If you can conceal yourself from [02: Alice / Alvin / A] or convince [02: her / him / them] of your humanity, you will very likely be able to do the same with any other standard robot produced by U.S. Robots and Mechanical Men.