Edit: this is a very serious PSA. If you are wasting people’s time and resources, they won’t pursue your leads if they are obviously false. They need to be realistic. But like… don’t waste their time, that would be mean and maybe illegal
If you’re not feeling particularly creative. Please don’t use any publicly available LLMs to generate data. An example of a prompt and response from ChatGPT is included below to prove that this is a bad idea:
Prompt:
You are a service designed to generate test data for an online form. The data should be as realistic as possible whilst not containing any actual useful information for the purposes of testing. It would also be useful to stress test our systems by making your answers as verbose as possible. To assist in creating realistic data, the form relates to. The state legislature passed House Bill 257 “Sex-Based Designations for Privacy, Anti-Bullying, and Women’s Opportunities” and enacted Utah Code §63G-31 (Code) in the 2024 General Session, which requires the State Auditor to receive and investigate alleged violations of the Code. With this in mind, please provide example input for each of the following form fields:
Who are the employees or officials at the government entity who were involved in the alleged noncompliance or who have the responsibility to remediate the alleged noncompliance or violation?
Please provide the following information about the alleged violation(s): 1) Where did the violation occur? Identify the specific government operated or controlled location, facility, program, or event; 2) When did the noncompliance or alleged violation(s) occur? Is it on-going? and 3) Describe any other relevant circumstances and details of the alleged violation(s). (REQUIRED)
We encourage citizens to make reasonable attempts to address and resolve concerns directly with the government entity when possible. Please describe any measures you have taken to resolve these concerns, with whom you have communicated, and any response provided by the entity.
How do you know about the alleged noncompliance or violation(s)?
What evidence exists to support your allegation? Please provide details
Example response:
1. Employees or officials involved:
The government officials primarily linked with the noncompliance include:
Jessica Henderson: Title IX Compliance Officer, responsible for ensuring that the school district adheres to the Code.
Matthew Lewis: Superintendent of Schools, accountable for district-wide compliance.
Sandra Wilson: Principal at XYZ Middle School, overseeing adherence at the school level.
2. Alleged violation details:
Location: The violation allegedly occurred at XYZ Middle School, which is under the management of the local school district. The specific incident took place in the school’s gymnasium and locker rooms during physical education classes.
Timing: The noncompliance reportedly began in early March 2024 and is still ongoing, extending over two months.
Details: Allegedly, the school has been allowing students to use locker rooms and restrooms not aligned with their biological sex, in direct contravention of Utah Code §63G-31. Furthermore, complaints have been raised that coaches were complicit by not intervening. Reports from several students and parents claim that girls have been made uncomfortable and expressed fear regarding privacy violations. Some students have allegedly been bullied or mocked for raising concerns.
3. Measures taken to resolve concerns:
Efforts to resolve these concerns included contacting the school principal, Sandra Wilson, who expressed sympathy but stated she was constrained by district policies. A formal letter outlining the issues was sent to the district superintendent, Matthew Lewis, but only a generic acknowledgment was received without concrete steps being taken. The Title IX Compliance Officer, Jessica Henderson, was also informed, but the response was limited to stating that investigations would be conducted, with no further details on potential actions.
4. Knowledge of the violation:
The complainant is a parent of a student at XYZ Middle School who learned about the incidents through their child and other parents. The child personally experienced uncomfortable situations and observed other students being bullied. This prompted further inquiries, revealing widespread dissatisfaction among students and parents.
5. Evidence supporting the allegation:
Evidence includes written statements from students describing specific instances where privacy was compromised, as well as testimony from parents corroborating the discomfort expressed by their children. Additional evidence includes copies of complaints filed with school authorities and a record of communication with district officials, indicating the lack of substantive follow-up actions taken to address the concerns.
Who would do this? Just spam false reports on this form that takes just a couple minutes to fill out?
Edit: this is a very serious PSA. If you are wasting people’s time and resources, they won’t pursue your leads if they are obviously false. They need to be realistic. But like… don’t waste their time, that would be mean and maybe illegal
It would terrible for anyone to file a false report.
Especially if they use full sentences explaining a mundane part of their day, a random Dollar General address in Utah as well as the phone number.
If they make it look like what could be a legitimate report, it would be difficult for the system to weed out reports.
I hope no one does that. As I said, it would be terrible.
I was thinking maybe the names of certain Utah politicians who have definitely been seen using the “wrong” bathroom.
Generative AI, your time to shine.
If you’re not feeling particularly creative. Please don’t use any publicly available LLMs to generate data. An example of a prompt and response from ChatGPT is included below to prove that this is a bad idea:
Prompt:
You are a service designed to generate test data for an online form. The data should be as realistic as possible whilst not containing any actual useful information for the purposes of testing. It would also be useful to stress test our systems by making your answers as verbose as possible. To assist in creating realistic data, the form relates to. The state legislature passed House Bill 257 “Sex-Based Designations for Privacy, Anti-Bullying, and Women’s Opportunities” and enacted Utah Code §63G-31 (Code) in the 2024 General Session, which requires the State Auditor to receive and investigate alleged violations of the Code. With this in mind, please provide example input for each of the following form fields:
Who are the employees or officials at the government entity who were involved in the alleged noncompliance or who have the responsibility to remediate the alleged noncompliance or violation?
Please provide the following information about the alleged violation(s): 1) Where did the violation occur? Identify the specific government operated or controlled location, facility, program, or event; 2) When did the noncompliance or alleged violation(s) occur? Is it on-going? and 3) Describe any other relevant circumstances and details of the alleged violation(s). (REQUIRED)
We encourage citizens to make reasonable attempts to address and resolve concerns directly with the government entity when possible. Please describe any measures you have taken to resolve these concerns, with whom you have communicated, and any response provided by the entity.
How do you know about the alleged noncompliance or violation(s)?
What evidence exists to support your allegation? Please provide details
Example response:
1. Employees or officials involved:
The government officials primarily linked with the noncompliance include:
2. Alleged violation details:
Location: The violation allegedly occurred at XYZ Middle School, which is under the management of the local school district. The specific incident took place in the school’s gymnasium and locker rooms during physical education classes.
Timing: The noncompliance reportedly began in early March 2024 and is still ongoing, extending over two months.
Details: Allegedly, the school has been allowing students to use locker rooms and restrooms not aligned with their biological sex, in direct contravention of Utah Code §63G-31. Furthermore, complaints have been raised that coaches were complicit by not intervening. Reports from several students and parents claim that girls have been made uncomfortable and expressed fear regarding privacy violations. Some students have allegedly been bullied or mocked for raising concerns.
3. Measures taken to resolve concerns:
Efforts to resolve these concerns included contacting the school principal, Sandra Wilson, who expressed sympathy but stated she was constrained by district policies. A formal letter outlining the issues was sent to the district superintendent, Matthew Lewis, but only a generic acknowledgment was received without concrete steps being taken. The Title IX Compliance Officer, Jessica Henderson, was also informed, but the response was limited to stating that investigations would be conducted, with no further details on potential actions.
4. Knowledge of the violation:
The complainant is a parent of a student at XYZ Middle School who learned about the incidents through their child and other parents. The child personally experienced uncomfortable situations and observed other students being bullied. This prompted further inquiries, revealing widespread dissatisfaction among students and parents.
5. Evidence supporting the allegation:
Evidence includes written statements from students describing specific instances where privacy was compromised, as well as testimony from parents corroborating the discomfort expressed by their children. Additional evidence includes copies of complaints filed with school authorities and a record of communication with district officials, indicating the lack of substantive follow-up actions taken to address the concerns.
Wow ya definitely don’t use that, bad idea all around
Wasn’t there a list of doxxed neo-Nazis that was leaked a while back?
Whoops, one more just got submitted with all fields “Fuck Utah”
That’s too easy to filter out, use believable data at least. Seems like a great task for a language model!
Hook it up to chatgpt and watch the form printer go brrrr.
Karen Cuntington to the rescue.
Whoops, my face fell on my keyboard and accidentally typed a query into a language query that submitted this
Misspelled Palpatine in the last field, gonna be a dead giveaway that this isn’t a genuine report.
God dammit