Social capabilities across human development | Part 14
Trust concerns actions after a promise, as well as the words. When plans become impossible, someone can leave another person waiting without information or explain the effect and discuss alternatives. Social capability also appears in adjustments after failure.
Running late for a meeting
On the way to meet a friend, a person discovers that the planned route is blocked. They contact the friend early and give an arrival estimate within what they can reasonably know. The waiting friend checks their plans and chooses to have tea first. On arrival, the delayed person checks the impact and apologizes.
This is fictional. Arriving on time was not achieved, but the other person received information for using their own time. An uncertain arrival should not be stated as certain; changes can be communicated again.
How children judge promises
Research on selective trust in promises examines children's use of factors such as another person's past behavior when judging a promise. It considers who acted in what way alongside the words used. See Children's Selective Trust in Promises.
An experimental trust choice cannot be generalized to every relationship. This article considers how trust experiences begin early and how different responsibilities arise in home, school, and work. It does not propose an age at which understanding promises is complete.
Promises change other people's plans
"I'll give it to you tomorrow" can shape the recipient's next task. Failure affects others, so distinguish notification, checking the impact, and agreement on new conditions.
Kindness does not require promising everything. Stating a limit beforehand preserves choices when success is uncertain. Changing a deadline also requires checking acceptance rather than merely announcing a new date.
Avoiding a permanent character judgment from one failure can coexist with acknowledging repeated harm. Look at changed actions when assessing repair.
A design hypothesis for physical AI
Before generating a promise, a possible design checks feasibility, deadlines, responsibility, and external dependencies. Agreed commitments should reference the same information as scheduling and motion plans and require notification when conditions change.
Do not generate "I can definitely do it" merely to increase approval. When a delay is known, explain its reason and what can be stated with certainty, then seek renewed agreement. Natural-sounding apologies should not automatically mark an unfulfilled promise complete.
Evaluation could include exceeded deadlines, external service interruption, cancellation, and rejection of changed conditions. Proposed measures include notification delay, unagreed deadline changes, and repeated failures.
A decision in four panels
- A friend waits at the meeting place.
- The delayed person discovers an obstacle and sends an update.
- The waiting friend uses their time after receiving it.
- They meet and check the impact.
The fictional comic does not guarantee that an apology restores trust. It focuses on information enabling the other person's judgment.
References
About this series
This series uses published literature to examine everyday situations. Its 7 domains and 22 capabilities are an editorial framework, not a standard academic classification. Differences in age or culture are not rankings of people's social worth or capability. The proposed AI designs do not establish humanlike emotional experience or demonstrated product performance. The comics illustrate fictional examples.
LOGIHEART's research and development vision
LOGIHEART is researching how to implement social capabilities in physical AI that interacts with people.
Explore the research and development vision at the LOGIHEART website and in the English introduction video.
Top comments (0)