CSE Roadmap

Common tier · about 4 hours · 12 parts

Part 9 of 12 · 2 min read

Mistakes that erode trust in the interface

Common mistakes, and what the example does not do.

Mistakes that erode trust in the interface

  • Streaming every internal event to the user.
  • Hiding errors or approvals inside a collapsed log.
  • Asking for approval on everything, then on nothing.
  • No undo, or an undo that silently fails.
  • A progress bar that is not tied to real steps.
  • Confident wording with no source or calibration.
  • Many questions at once, or questions the user already answered.
  • No way to stop a long run.
  • Starting over after a failure instead of resuming.
  • Not testing with real users until the end.

What the example does not do

About this example What the example does not do
  • No real UI: it is a view model and policy logic; layout, accessibility, animation and mobile behaviour are not covered.
  • The fatigue model is invented; real attention depends on stakes, wording, training and context, so the percentages are only illustrative.
  • The ambiguity model is invented: real models' scores are not Gaussian noise around fixed means.
  • Event streams are random, not recorded from a real agent.
  • No user testing: the example measured items shown and invariants, not comprehension, trust or time to notice an error.
  • The undo log is in memory, with no persistence, concurrency or third-party side effects.
  • Optional real experiment: show five users a real agent run in the raw log and in the layered view, and plant one risky approval. Then measure completion time, whether they caught the risky item, and how they rate trust.

Check yourself

A run produces 82 events. What does the person see by default?

Show answer

Not 82 items: a grouped timeline showed 26 and the summary view 1.4, while every unresolved approval and unrecovered error stayed visible. That invariant is tested.

For the curious Fewer prompts protect the dangerous ones

Assume an attention model where the chance of reading a prompt falls 7 percent with each one answered. Asking about all 20 actions led to only 52 percent of the 2 risky actions being read. Tiered prompting reached 79 percent, and tiered and batched reached 86 percent. Asking when the margin between readings is under 0.2 cut wrong actions from 13.3 to 1.5 percent at 44 percent asks. The parameters are assumed; the lesson is general.

What you will use daily at work

Practice What it looks like
Answer three questions on screen What is it doing, what needs you, how do you stop or undo it.
Tier and batch approvals Unanswered means denied.
Give an undo or delay for outbound actions List what cannot be undone.
Test with five or so real users Before polishing.