Video search readers are already asking about coverage
By OnTheRice · Sources and limitations · Report a correction
A fresh direct read of the Show HN thread shows readers testing the cost and completeness of searching video frames, not only admiring the demo. One commenter says sampling one frame per second across 12,000 videos took days on an Apple M1; keyframes took overnight. Others ask whether brief moments could be missed and how much work large libraries need. The current page showed 62 points and 37 comments, one HN snapshot. The older saved indexed reading conflicts with the current page’s relative age, so no growth claim is made. Next check: request benchmarked time and recall.
Source confidenceNot assessed
What’s changing?
A visible HN discussion moves from feature novelty toward compute time, sampling coverage and search usefulness on large private media libraries.
Key points
- At 2026-10-04T16:43:40Z, the direct HN item for the Show HN photo/video search project displayed 62 points, 37 comments and an age of 6 hours. Visible comments include one user's report that sampling one frame per second from 12,000 videos took days on an Apple M1 and keyframes took overnight; other comments raise missed brief appearances and large-library processing time. An earlier saved HN index observation for this same object, captured at 2026-10-04T15:04:51Z, displayed 6 points and 1 comment and described an age of about 2 hours. The current page's relative age conflicts with that earlier display, so the counters are not used as a growth series.
What could change this view?
- Some commenters praise the ability to jump to a video moment, and one recommends proxy workflows; the discussion contains both use cases and friction.
Bottom line
Those limits could decide whether users try the tool or trust that a short moment will be found, but current comments do not establish product performance.
This is based on the information recorded with this signal.
How likely is this call? · Three frameworks
This signal has no investible bullish or bearish call. A market-direction percentage does not apply.
Evidence behind this view
Checks whether separate pieces of evidence support the same view.
Recorded connection
A visible HN discussion moves from feature novelty toward compute time, sampling coverage and search usefulness on large private media libraries.
We have not established a separate count of independent evidence clusters for this card. Several reports may rely on the same event or data provider.
What past price moves show
Compares today’s price moves with recorded past patterns.
This card has no recorded investible call. A price-pattern comparison is not applicable unless a specific asset is added with evidence.
Evidence against this view
Shows evidence and risks that could challenge this view.
Recorded opposing evidence and risks
Some commenters praise the ability to jump to a video moment, and one recommends proxy workflows; the discussion contains both use cases and friction.
Why the direction changed
Explains recorded changes from expecting a rise to expecting a fall, or the reverse.
No documented bullish-to-bearish or bearish-to-bullish change is available for this card. We keep the original call and its time unchanged.
What people say vs the facts
Compares recorded public discussion with checked market facts.
A fresh direct read of the Show HN thread shows readers testing the cost and completeness of searching video frames, not only admiring the demo. One commenter says sampling one frame per second across 12,000 videos took days on an Apple M1; keyframes took overnight. Others ask whether brief moments could be missed and how much work large libraries need. The current page showed 62 points and 37 comments, one HN snapshot. The older saved indexed reading conflicts with the current page’s relative age, so no growth claim is made. Next check: request benchmarked time and recall.
This records the public discussion described by the signal. No separate comparison between that discussion and verified market prices is recorded here. It is not a measure of wider public opinion.
How past forecasts performed
Shows recorded forecast results and what they can teach us.
This card has no recorded investible forecast to score.
No tested mapping from evidence combinations to future success is recorded for this card. Confidence in sources and past outcomes do not establish a calibrated probability.
Sources, original wording and price checks
Published
First published through Sites 05 Oct 2026, 01:54:01 SGT
No explicit investible BULLISH or BEARISH call on this card.
What changed
Recorded public historyA fresh direct read of the Show HN thread shows readers testing the cost and completeness of searching video frames, not only admiring the…
This is the first recorded public view. Further public updates will appear here.
- Video search readers are already asking about coverage
A fresh direct read of the Show HN thread shows readers testing the cost and completeness of searching video frames, not only admiring the demo. One commenter says sampling one frame per second across 12,000 videos took days on an Apple M1; keyframes took overnight. Others ask whether brief moments could be missed and how much work large libraries need. The current page showed 62 points and 37 comments, one HN snapshot. The older saved indexed reading conflicts with the current page’s relative age, so no growth claim is made. Next check: request benchmarked time and recall.
First published by OTR: 5 Oct, 01:54 SGTPublished by OTR: 5 Oct 2026, 01:54 SGT
Why the view changed
This is the first recorded view. Future source-backed updates will appear here.
Evidence, memory & review 2 linked sources
Directional takeaway: N/A: No listed-asset exposure or asset-specific investment direction is supported.
OTR reasoning: A visible HN discussion moves from feature novelty toward compute time, sampling coverage and search usefulness on large private media libraries.
Counterpoint: Some commenters praise the ability to jump to a video moment, and one recommends proxy workflows; the discussion contains both use cases and friction.
Source-grounded facts (paraphrased)
- Time not supplied: At 2026-10-04T16:43:40Z, the direct HN item for the Show HN photo/video search project displayed 62 points, 37 comments and an age of 6 hours. Visible comments include one user's report that sampling one frame per second from 12,000 videos took days on an Apple M1 and keyframes took overnight; other comments raise missed brief appearances and large-library processing time. An earlier saved HN index observation for this same object, captured at 2026-10-04T15:04:51Z, displayed 6 points and 1 comment and described an age of about 2 hours. The current page's relative age conflicts with that earlier display, so the counters are not used as a growth series.
Source trail
- Show HN: AI search for every photo and every frame of video on macOS ↗Hacker NewsSource time not supplied
- AI search for every photo and every frame of video on macOS ↗GitHub / project ownerNo publication clock recorded in this pass · not stated
Review status: SUPPORTED WITH CAVEATS
No score or probability is presented.