为什么我越来越在意普通人的问题Why everyday problems matter more and more to me
从适老手机、女性健康到家庭照护：那些不宏大、却反复消耗人的问题，为什么值得被认真设计。From age-friendly phone tools to women's health and family care, I keep returning to problems that draw little attention but wear people down over time.
教一个不熟悉智能手机的人完成操作时，我们很容易说：“点右上角，再进设置，把权限打开就行。”
对熟悉界面的人来说，这是一条再普通不过的路径。但对另一个人来说，其中每一步都可能带着不确定：哪个才是右上角的按钮？点错会不会扣钱？弹出的提示是什么意思？如果退出了，还能不能回到刚才的地方？
真正让人停下来的，往往不是某一个按钮，而是连续做出判断的压力。
我后来发现，自己反复想做的产品，似乎都在面对这样的时刻。
EasyPhone AI 想处理的是长辈与智能手机之间的门槛；“她也这样”关心身体和情绪的长期变化如何被记录；“爸妈今天还好吗”试着让异地家庭的关心发生在日常，而不是只在异常之后出现。它们表面上属于不同领域，背后却有一条相同的线：一些看起来不够宏大、难以成为新闻，却每天都在消耗人的问题。
普通，不等于不重要
我们习惯用规模证明一个问题值得解决：影响了多少人，市场有多大，效率能提升多少。这样的判断当然必要，但它也容易漏掉另一类问题。
它们单次造成的损失不大，却会反复出现；不一定让人立刻求助，却会长期积累挫败感；很难用一个漂亮数字概括，却真实地影响着一个人能否保有自主、体面和与他人的连接。
不会使用某个功能，并不只是少完成了一次操作。一个人可能因此越来越不愿意碰手机，越来越依赖家人，也越来越少参与已经迁移到数字世界里的生活。
一段身体变化没有被记录，也不一定马上导致严重后果。但当睡眠、疼痛、情绪和用药都只能依靠模糊的回忆，人与自己的身体之间就少了一份可以回看的线索。
异地的子女没有每天打电话，并不代表不关心父母。很多时候，真正缺少的是一种负担更小、不过度打扰、又能让关心持续发生的方式。
这些都不是“改变世界”的宏大命题。但如果产品愿意认真对待其中一个人，世界对那个人而言，确实可能变得容易一点。
温柔不是一种配色
做这类产品时，“温暖”“陪伴”“关怀”是很容易被使用的词。界面换成柔和的颜色，文案多说几句“别担心”，好像就已经完成人本设计。
但我越来越觉得，真正的温柔首先是一种系统行为。
它意味着少让用户做一次不必要的选择；在执行前确认，而不是自作主张；在风险变高时停下来，而不是为了流程完整继续给答案；不把照护变成全天候监控；不把情绪波动包装成诊断；也不让“为了你好”成为读取更多隐私的理由。
以面向长辈的手机助手为例，语音入口和大字界面只是表层。更重要的是，系统能不能听懂一个不完整的表达，能不能一次只说清一步，能不能在转账、验证码、陌生链接等高风险场景中明确中断，能不能在自己做不到时，生成一张简单的求助卡，把问题交给可信任的家人。
产品的关怀，不应该只出现在它顺利工作的时候。失败时怎样保护用户，才更接近它真正的价值观。
AI 不替代关系，只帮助人适时出现
当 AI 进入健康和家庭场景，最诱人的叙事之一，是“它可以一直陪着你”。但我对此始终有一点警惕。
技术可以提醒、整理、降低表达成本，也可以帮助人看见原本零散的变化。它却不应该假装成亲情本身，更不该用“陪伴”掩盖真实关系的缺席。
我更认可的方向是：AI 不替代家人，而是帮助家人在合适的时刻出现。
它可以把几天的记录整理成一段简短摘要，让一次电话不再从“最近怎么样”开始又很快结束；可以在长辈遇到复杂操作时，把求助变得更具体；可以提示一段变化值得关注，但把判断和行动留给本人、家人或专业人士。
这里面最难的不是让 AI 多做一点，而是决定它应该在哪里停下。
从“用户不会”转向“产品是否足够理解人”
过去遇到使用障碍时，我们常把原因归结为用户不会、记不住、不愿学。但如果同一种困惑持续发生，也许更该被追问的是：产品为什么要求每个人都用同一种方式理解它？
年龄、教育经历、身体状态和数字经验不同，人面对技术时拥有的耐心、信心和判断成本也不同。所谓“边缘用户”，很多时候只是主流设计没有认真看见的人。
这也是我越来越在意普通人问题的原因。它们迫使我离开功能清单，去理解一个人在什么情境下打开产品、为什么犹豫、害怕什么、愿意付出多少注意力，以及即使完成任务之后，又承担了什么新的成本。
我还不知道这些探索最终会长成一个产品、一组服务，还是一条更长期的工作方向。现阶段的很多东西也只是原型，距离真实可用仍有很长的验证过程。
但我愿意先把这些问题留下来。
不是每个产品都要改变世界。至少，它可以认真对待一个人的为难；至少，技术不必总让适应它的人独自承担代价。
如果长期变化值得被温柔看见，那么那些长期被忽略的人，也一样值得。
Teaching someone unfamiliar with smartphones, it’s easy to say, “Tap the top right, go into settings, and turn on the permission.”
For someone who knows the interface, that’s a routine sequence. For someone else, each step can bring uncertainty. Which button at the top right? Could a wrong tap cost money? What does that pop-up mean? If I leave this screen, can I get back?
The strain of making one decision after another can stop someone before any single button does.
I began to notice that the products I kept wanting to build dealt with moments like these.
With EasyPhone AI, I want to address the barriers older adults face with smartphones. With “她也这样” (She Feels It Too), I am exploring ways to record changes in the body and mood over time. With “爸妈今天还好吗” (How Are Mum and Dad Today?), I am trying to help families living apart express care in daily life, before something goes wrong. These ideas sit in different fields, but I keep finding the same concern in them: problems that rarely make the news yet cost people something every day.
Ordinary problems deserve attention
We tend to justify solving a problem through scale: how many people it affects, how large the market is, how much efficiency we could gain. Those judgments matter. They can also leave some problems out.
One occurrence may cost little, but repeated difficulties add up. A person might not ask for help at once, yet carry a growing sense of frustration. A neat metric may miss the effect on their independence, dignity, and connection with others.
Failing to use a phone feature can mean more than one unfinished task. Someone may become less willing to touch their phone, more dependent on family, and less able to take part in activities that have moved online.
Not recording a physical change may have no immediate serious consequence. But relying on hazy memories of sleep, pain, mood, and medication leaves a person with less to look back on when trying to understand their body.
Adult children living far away may care about their parents without calling every day. They may need a less demanding, less intrusive way to stay involved over time.
These are modest problems compared with the ambition to “change the world.” Taking one person’s difficulty seriously could still make their daily life easier.
Kindness takes more than a color palette
Words like warmth, companionship, and care come easily in this kind of product work. It can feel as if softer colors and a few reassuring messages are enough to call a design human-centered.
I have come to think that care starts with how the system behaves.
I want to spare users unnecessary choices and ask before taking action. I want the system to stop when risk rises, even if that interrupts the flow. Care should not become round-the-clock surveillance. Mood changes should not turn into diagnoses. “For your own good” should not excuse reading more private information.
A voice input and large text are a start for a phone assistant aimed at older adults. I care more about whether it can handle an incomplete explanation, give one clear step at a time, and stop at risky moments involving money transfers, verification codes, or unfamiliar links. If it cannot help, can it prepare a simple request for help that the user can pass to a trusted family member?
I want users to feel that care when the product fails, too. The protection they receive at that moment says a great deal about its values.
AI can help people show up for each other
In health and family settings, the promise that AI “can always be there for you” is tempting. I remain cautious about it.
We can use technology to send reminders, organize information, reduce the effort of expressing a need, and notice changes across scattered records. We should not pretend it can be family, or use the language of companionship to cover up the absence of human relationships.
I would rather help family members be there at the right moment.
AI could summarize a few days of records so a call has somewhere to go beyond “How have you been?” It could make a request for help more specific when an older person gets stuck on their phone. It could flag a change worth attention while leaving judgment and action to the person, their family, or a professional.
The hardest part is deciding where the AI should stop.
I want to question the product before blaming the user
Faced with a usability problem, we often say that users don’t know how, can’t remember, or won’t learn. If the same confusion keeps recurring, I want to ask why the product expects everyone to understand it in the same way.
Age, education, physical condition, and digital experience affect how much patience and confidence someone brings to technology, and how much effort each decision takes. The people we call “edge cases” may be people mainstream designers haven’t taken time to understand.
These everyday problems pull me beyond feature lists. I need to understand the situation in which someone opens a product, why they hesitate, what they fear, and how much attention they can give. I also need to ask what new costs they carry after completing the task.
I don’t know whether these explorations will grow into one product, a set of services, or a longer-term direction for my work. Much of what I have now is still at the prototype stage, with substantial testing ahead before I can call it useful in practice.
For now, I want to keep working with the questions.
A product can take one person’s difficulty seriously without claiming to change the world. People adapting to technology should not have to bear all the costs of that adjustment alone.
I want to give patient attention to changes that unfold over time, and to people whose difficulties we have overlooked for too long.