跳到正文
Hacker News· kbrower·· 6 天前AI 评分38

Health Loom:把健康疑问变成动画解说视频

Turn health questions into explainer videos

AI 导读

Health Loom 将听众提交的健康问题转化为带旁白和手绘动画的解说视频,每个数字都追溯到原始研究,提问与观看均免费。其 AI 追问环节由 Anthropic 的 Claude 模型负责,只收集背景信息、不提供诊断或建议,用户默认匿名。用户还可选择将匿名健康背景数据捐入研究数据集,用于未来的流行病学研究。

正文 · 原文

Your health question, woven into an animated explainer.

Send us the health question you keep wondering about. We dig into the research, check every number against its source, and turn the answer into a short, narrated video anyone can follow.

Free to ask. About five minutes. Anonymous unless you'd like a shout-out.

How it works

From a question in your head to a video on your screen

Every Health Loom episode starts with a real listener question. Here's what happens to yours.

  1. You ask

    Write your question and answer a few optional background questions. Our AI interviewer asks a few skippable follow-ups, and you pick a video length, from a quick answer to a podcast. Right now, every question becomes a video.

  2. We research

    We go to the primary studies and guidelines, grade the evidence honestly, and note where it's thin. Every number in the script traces to a source.

  3. We animate

    The answer becomes a plain-language script, a narrated voice track, and a hand-drawn animation, with captions and chapters.

  4. You watch

    Leave an email and we'll tell you when your episode is live. We read the question on air without your name, unless you ask for a shout-out.

An example episode

“Are You Taking Your Blood Pressure Wrong?”

What are all the ways people take their blood pressure incorrectly, how much does each one push the reading up or down, and what should you do if taking it makes you anxious?

About ten minutes, in seven chapters:

  1. 0:00Intro
  2. 0:59The cuff
  3. 2:17How you sit
  4. 3:11What you did beforehand
  5. 4:33During the reading
  6. 5:48When the cuff makes you anxious
  7. 8:47Before you go

The sources run from a 1983 Lancet study of the doctor effect to a 2025 randomized trial on noisy rooms, and every one is listed in the episode description.

Open on YouTube

Our editorial standards

Evidence first, plain language always

Primary sources

We read the studies, not the headlines about them, and list every source in the video description.

Honest about certainty

A big trial and a small survey aren't equal. We say how strong the evidence is, and when it's mixed.

Plain spoken

No jargon without a translation. If a term matters, we explain it once, clearly, then use it.

Education, not advice

Our videos explain what the evidence says in general. Decisions about your own care belong with your clinician.

Optional: help future research

Your question is a data point, too

The questions people ask, and the lives behind them, say a lot about what's affecting health right now. Every question comes with a short, optional health background, which helps us tailor the answer to people like you.

If you agree, an anonymous copy of that background also goes into our research dataset. Over time it could become a primary source for epidemiology: what people are worried about, who's asking, and how that lines up with conditions and habits. Saying no doesn't affect your question.

  • Opt-in only. You choose whether a research copy is made, and you can skip any question.
  • Stored apart. The research copy is kept separately from your question, name and email.
  • Coarse by design. Age ranges, not birthdays, and never where you live.
  • Ethics review first. We'll only use it in a study after independent ethics review, and only share it in de-identified or aggregate form.
  • Leave anytime. You get a withdrawal code; send it to us and we'll delete your answers.

Questions about questions

FAQ

Is this medical advice?

No. Health Loom explains what the evidence says for people in general. We can't diagnose you or recommend treatment, so please talk to a clinician about your own situation. If you might be having an emergency, call 911 or your local emergency number now.

Will you make a video about my question?

Yes. Right now we make a video for every listener question we receive. The more detail you give, the better it can answer your situation. If you leave an email, we'll tell you when we start on yours and send the link when it's live. If questions ever outpace what we can make, we'll say so here.

Can I stay anonymous?

Yes, that's the default. We paraphrase questions on air and take out anything that could identify you. If you'd like a shout-out, you can tell us how to credit you.

Why does an AI ask me follow-up questions?

A question like "Is salt bad for me?" could lead to five different videos. A few quick follow-ups, plus a short last round for anything still unclear, tell us which one you need. The interviewer is an AI model (Claude, made by Anthropic). It doesn't answer your question or give advice; it only asks for context, and you can skip any follow-up.

Why do you ask about my health history?

So the follow-ups and the video fit your situation: a question about salt means something different for someone with kidney disease. Every item is optional. On air we may mention details you share, like your age, conditions or medicines, but never where you live, and you're named only if you choose to be credited. Separately, you can choose to add an anonymous copy to our research dataset. See our privacy and research notice for the details.

How are the videos made?

With people and AI tools working together. AI helps with the follow-up interview, research and drafting, and the narration is an AI voice. Every claim is checked against the primary sources, which are listed in each video's description.

Does it cost anything?

No. Asking is free, and so are the videos.

What have you always wanted to ask?

Big or small, common or strange. If you're wondering, someone else is too.

Ask your question

来源:Hacker News · healthloom.com