HN.watch presents the Hacker News front page with a short AI-generated video explainer attached to every story. The site's front page describes itself as "Hacker News with explainer videos" and states: "Click a title to watch." The current feed lists about 30 stories, each marked with a video badge alongside standard metadata (points, comment count, relative time).
The article describes HN.watch as 'Hacker News with explainer videos,' where each of the roughly 30 front-page stories carries a video badge alongside standard metadata.
The service was submitted to Hacker News earlier this week and received 161 points. A supporting discussion thread on a separate story covered by HN.watch (Phyllotaxis: An audio-reactive LED display) provides some user feedback on the AI explainers. Users reported that the videos can strip away nuance from articles and state interpretations as fact without naming the source. One commenter gave a concrete example: an article quoted a subject saying "That gunshot prematurely aged my brain," while the AI summary rendered it as "The impact prematurely aged her brain," presenting the claim without the original framing or attribution.
Other feedback noted that the AI voices sound monotonous and that the videos can feel repetitive.
The primary source (hn.watch) does not disclose the underlying AI models, the video generation pipeline, or the cost per video. The site's footer lists standard links (Guidelines, FAQ, Lists, API, Security, Legal, Apply to YC, Contact) and sharing options (HN comment link, X/Twitter).
What remains uncertain: whether the misrepresentation issue is systematic or occasional, which AI models power the video generation, and whether the site offers any editorial controls beyond clicking a story title.