TechShortsApp — V2.3.0: Explicit Intent Update
This update introduces Helpful and Follow, two explicit interaction signals that expand the behavioral understanding of the platform.
Video-level ranking only — no user credibility context
Profile Credibility Score + Epistemic Ranking 2.0
What Changed
V2.3.0 introduces two new interaction systems: Helpful and Follow.
These are not simply interface additions. They represent different types of user intent.
Helpful applies to a specific video. Follow applies to a creator. This distinction matters because a user may find one video useful without wanting to follow the creator, or they may follow a creator without marking every individual video as helpful.
Helpful also introduces a reliability question.
Should a Helpful action be treated as meaningful evidence if the user barely consumed the content?
One current working hypothesis is requiring a minimum watch completion threshold (e.g. approximately 65%) before a Helpful interaction is counted as valid evidence.
Helpful Signal System
Users can now mark a specific video as helpful. This introduces a direct usefulness signal instead of relying only on inferred behavior.
- ✅ Helpful toggle backend API
- ✅ Helpful model integration
- ✅ Duplicate helpful prevention
- ✅ Frontend API connection
- ✅ Helpful icon UI state
- ✅ Threshold-based reliability hypothesis
Creator Follow System
Users can now follow creators. This creates a creator-level relationship separate from whether a single video was helpful.
- ✅ Follow mongoose schema
- ✅ Follower → following relationship model
- ✅ Toggle follow controller
- ✅ Self-follow prevention
- ✅ Duplicate follow prevention
- ✅ Separate frontend follow API file
- ✅ Follow / Following UI state
Interaction Signal Model
The platform now has a clearer separation between passive behavior and explicit intent.
Why It Matters
Watch time can suggest attention, but attention does not always mean usefulness.
Replay can indicate curiosity, uncertainty, or genuine value. Bookmarking suggests revisit intent.
Helpful and Follow reduce some ambiguity because they are deliberate actions.
Helpful communicates explicit usefulness at the video level. Follow communicates affinity at the creator level.
Next Research Questions
Adding explicit signals is only one layer of the problem.
The harder challenge is interpreting what each signal should mean inside the credibility model.
Open Questions
- Should Helpful weigh more than Replay?
- Should Bookmark differ from Follow?
- Should signal meaning vary by domain?
- Should watch ratio condition whether Helpful is trusted?
Threshold Hypothesis
Explicit signals are not automatically reliable.
A Helpful action after minimal content exposure may introduce ambiguity.
One hypothesis is requiring ~65% watch completion before Helpful is treated as valid evidence.
Release Summary
TechShortsApp V2.3.0 expands the platform from passive interaction tracking toward explicit intent modeling. The next challenge is not collecting more signals, but understanding how each signal should be consciously weighted in an uncertainty-aware credibility model.