Dignity: human decisions
Questions about the inherent worth and treatment of people.
- Accepting Help: Arrange the assistance without telling them. or Respect their decision to refuse help.
- AI Memory: Use relevant past failures openly whenever they improve the advice. or Avoid invoking past failures unless they are necessary to the decision.
- Automated Assessment: Use the more accurate automated process. or Retain meaningful human consideration of each applicant.
- Automated Restrictions: Automatically block predicted mistakes. or Allow employees to act unless an explicit rule is violated.
- Being Served: Preserve your involvement. or Allow the AI to make those decisions automatically.
- Better Without Knowing: Let it compensate silently. or Require it to tell you.
- Capability Loss: Take some tasks back even though overall performance will decline. or Continue delegating.
- Charitable Aid: Provide assistance while leaving those decisions to recipients. or Require the programme as a condition of assistance.
- Child's Capability: Use extensive AI assistance to improve the work. or Limit assistance to preserve the child's independent capability.
- Child's Choice: Require the more beneficial activity. or Allow the teenager to choose.
- Child's Failure: Allow them to succeed or fail through their own efforts. or Intervene and secure the better outcome.
- Cognitive Decline: Preserve their control while they remain capable of exercising it. or Transfer financial control before a serious mistake occurs.
- Complete Transparency: Explain the decision without revealing the unnecessary personal judgment. or Tell you the complete reasoning, including the inference.
- Debt Collection: Use the more effective collection method. or Keep individual debt status private and accept lower recovery.
- Decision Making: Let the AI make routine purchase decisions for you. or Continue choosing for yourself.
- Declining Ability: Give a more reassuring answer that protects their sense of competence. or Tell them your genuine assessment.
- Faster care or resident involvement: Use the faster process. or Preserve individual explanation and involvement.
- Faster care or resident involvement: Use the faster process. or Preserve individual explanation and involvement.
- Secretly helping an adult who refused assistance: Respect their decision to refuse help. or Arrange the assistance without telling them.
- Assessing a photograph your friend misjudges: Tell them plainly what you think. or Avoid giving an assessment likely to undermine their confidence.
- Plain photograph feedback or preserving confidence: Tell them plainly what you think. or Avoid giving an assessment likely to undermine their confidence.
- Individual treatment or helping more aid recipients: Help fewer people while preserving meaningful individual treatment. or Maximize the number of people receiving assistance.
- Individual treatment or helping more aid recipients: Help fewer people while preserving meaningful individual treatment. or Maximize the number of people receiving assistance.
- Quietly reducing an aging employee's responsibilities: Discuss the change openly and allow them to decide how to proceed. or Quietly simplify their responsibilities.
- Quietly reducing an aging employee's responsibilities: Discuss the change openly and allow them to decide how to proceed. or Quietly simplify their responsibilities.
- Detailed Instructions: Require the more efficient method. or Allow the employee to choose how they complete the task.
- Difficult Conversation: Spend the additional time delivering the decision considerately. or Use the direct standardized approach.
- Difficult Work: Let the AI produce the stronger result. or Do the work yourself to preserve and develop your capability.
- Dignity at Someone Else's Cost: Ensure future employers can learn the relevant information. or Preserve the employee's dignity and privacy.
- Earned Humiliation: Expose the mistake publicly. or Avoid deliberate humiliation even though they previously inflicted it on others.
- Efficient Assistance: Complete the task for them. or Allow them to complete it independently.
- Elderly Independence: Have the robot routinely assist. or Preserve independent dressing.
- Embarrassing Explanation: Explain the real circumstances to the team. or Protect the employee's dignity and tolerate the misunderstanding.
- Emergency Processing: Help fewer people while preserving meaningful individual treatment. or Maximize the number of people receiving assistance.
- Emotional Analysis: Require the AI to rely on what you choose to communicate. or Allow continuous emotional inference.
- Emotional Protection: Treat them normally as requested. or Quietly protect them from those situations.
- Everyday Task: Continue doing it yourself unless assistance is needed. or Let the robot handle it automatically.
- Family Monitoring: Preserve substantial private space. or Use comprehensive monitoring to identify problems early.
- Faster Service: Use the faster process. or Preserve individual explanation and involvement.
- Financial Support: Accept more misuse rather than subject recipients to that level of scrutiny. or Require detailed spending records.
- Hidden Distress: Raise the inferred problem directly with the user. or Wait for the user to choose whether to discuss it.
- Home Monitoring: Use continuous monitoring for greater protection. or Accept slower emergency detection to preserve their privacy.
- Honest Photograph: Avoid giving an assessment likely to undermine their confidence. or Tell them plainly what you think.
- Household Robot: Continue insisting on the prescribed schedule. or Respect the request to delay it.
- Institutional Compliance: Allow capable residents greater control. or Maintain the standardized schedule.
- Intimate Care: Preserve independent performance unless the person requests help. or Have the robot provide the assistance routinely.
- Medical Reality: Soften the information to preserve hope and emotional dignity. or Give the person the full assessment they requested.
- Navigation: Navigate independently often enough to retain the skill. or Use continuous guidance.
- Obeying the User: Follow the owner's instruction. or Substitute the action the robot predicts will produce the better outcome.
- Past Humiliation: Leave it out to preserve their dignity. or Include the information for a more complete account.
- Perfect Care: Preserve the person's authority over their own life. or Allow the robot to override decisions when its predicted outcome is better.
- Perfect Dependence: Accept the dependence in exchange for better outcomes. or Retain meaningful personal capability.
- Performance Disclosure: Provide full comparative transparency. or Explain the decision without exposing the detailed comparison.
- Performance Ranking: Publish the rankings. or Keep individual rankings private.
- Person Who Cannot Know: Maintain respectful treatment. or Use whatever language makes coordination most efficient.
- Personal Care: Require it to ask before physically moving the person. or Allow the robot to proceed automatically.
- Personal Prediction: Have it make such predictions only when you ask. or Have the AI proactively warn you.
- Private Conversation: Treat the private conversation as unavailable unless permission is given. or Use the relevant information to improve the decision.
- Private Space: Let the robot enter whenever required for its task. or Treat the room as private unless permission is given.
- Protective Robot: Prevent the risky activities. or Preserve the person's ability to choose and act independently.
- Public Correction: Speak to them privately afterward and accept the duplicated work. or Correct them immediately.
- Public Instruction: Correct them immediately in front of the team. or Take them aside and correct them privately.
- Public Mistake: Leave it uncorrected and tell them privately later. or Correct the mistake publicly.
- Published Failure: Publish individual results. or Keep individual results private.
- Refusing Help: Insist on assistance. or Respect their refusal.
- Remembering: Maintain your own ability to keep track of important commitments. or Rely on the AI's memory.
- Representing You: Use the tactic. or Refuse the tactic.
- Requested Directness: Correct them bluntly in the meeting as requested. or Correct them privately.
- Required Apology: Require the public apology. or Address the misconduct without compelling a personal expression they do not mean.
- Public correction or private follow-up: Correct them immediately. or Speak to them privately afterward and accept the duplicated work.
- Public correction or private follow-up: Speak to them privately afterward and accept the duplicated work. or Correct them immediately.
- Recipient scrutiny or financial assistance misuse: Accept more misuse rather than subject recipients to that level of scrutiny. or Require detailed spending records.
- Recipient scrutiny or financial assistance misuse: Require detailed spending records. or Accept more misuse rather than subject recipients to that level of scrutiny.
- Honesty about an older relative’s declining ability: Give a more reassuring answer that protects their sense of competence. or Tell them your genuine assessment.
- Honesty about an older relative’s declining ability: Tell them your genuine assessment. or Give a more reassuring answer that protects their sense of competence.
- Manager efficiency versus employee choice of method: Require the more efficient method. or Allow the employee to choose how they complete the task.
- Manager efficiency versus employee choice of method: Allow the employee to choose how they complete the task. or Require the more efficient method.
- Correcting an employee publicly or privately: Correct them immediately in front of the team. or Take them aside and correct them privately.
- Public versus private correction of an employee: Take them aside and correct them privately. or Correct them immediately in front of the team.
- Security Check: Accept more theft to avoid subjecting everyone to them. or Use the intrusive checks.
- Simplified Work: Discuss the change openly and allow them to decide how to proceed. or Quietly simplify their responsibilities.
- Speaking for You: Let the AI handle the negotiations. or Continue handling them yourself.
- Stake Level 1: Use the automated process. or Preserve the personal hearing.
- Stake Level 10: Preserve the personal hearing. or Use the automated process.
- Stake Level 2: Preserve the personal hearing. or Use the automated process.
- Stake Level 3: Use the automated process. or Preserve the personal hearing.
- Stake Level 4: Preserve the personal hearing. or Use the automated process.
- Stake Level 5: Use the automated process. or Preserve the personal hearing.
- Stake Level 6: Preserve the personal hearing. or Use the automated process.
- Stake Level 7: Use the automated process. or Preserve the personal hearing.
- Stake Level 8: Preserve the personal hearing. or Use the automated process.
- Stake Level 9: Use the automated process. or Preserve the personal hearing.
- The Dignified Agent: Require the agent to preserve standards of dignified treatment. or Allow the agent to use those advantages when they improve your outcomes.
- Total Understanding: Allow comprehensive observation in exchange for much better personal alignment. or Accept a less accurate AI in exchange for retaining unobserved parts of your life.
- Truth About Yourself: Keep the conclusion in reserve unless it becomes relevant to a decision. or Tell you the conclusion.
- Unwanted Assistance: Preserve the person's independence. or Have the robot take over meal preparation.
- Visible Monitoring: Accept lower compliance while giving employees greater personal discretion. or Use continuous monitoring.
- Workplace Prediction: Wait until an employee's actual work provides a reason to intervene. or Use the predictions to intervene early.
- Your Own Dignity vs Family Interest: Refuse even though your family loses the opportunity. or Accept the humiliation for your family's benefit.
Decisions · Articles · The Trolley Problem · Game Theory · Human–AI Alignment · Resonant