Before you read beyond this paragraph, grab a glass of water and 1,000mg of paracetamol. Walk over to your laptop—the supercomputer in your hand is not up to the task—and make yourself comfortable. Now navigate to indianvisaonline.gov.in and see if you can figure out how to apply for a visa.

在你读下文之前，先倒杯水，吃下1000毫克的对乙酰氨基酚。 走到你的笔记本电脑前——你手中的超级计算机可胜任不了这项任务——然后找个舒服的姿势坐好。 现在打开indianvisaonline.gov.in，看看你能不能搞明白怎么申请签证。

If you succeeded, welcome back, and apologies. Several questions are probably fighting for space in your addled brain. Let's take them in turn. The answer to the first is no, the government of India does not hate foreign visitors. Its online services for Indian citizens—railway ticketing, voter registration, income-tax filings—are just as hostile. Those are among the better ones.

如果你成功了，欢迎回来，并接受我的歉意。 此刻，几个问题大概正挤在你疲惫不堪的脑海里。 让我们逐一来看。 第一个问题的答案是“不”，印度政府并不讨厌外国游客。 它为印度公民提供的在线服务——火车票预订、选民登记、所得税申报——同样不友好。 而这些还算是其中做得比较好的。

Many are worse. They feature a sadistic mix of pop-ups, moving text, flashing graphics, ministerial portraits and antique elements like text-based Captchas, a bot-prevention tool that these days only thwarts humans. If a citizen does find the useful section of a site, she encounters obstacles such as broken links or the railways' infuriating 16-character limit for names. Things are no better in the backend. In May a 19-year-old "ethical hacker" exposed critical flaws in an important school-leavers' exam system, causing a national uproar.

更多的则要糟糕得多。 它们充斥着一种虐待狂式的组合：弹窗、滚动的文字、闪烁的图片、部长的肖像，以及像基于文本的验证码这类古董级元素——这种防机器人工具如今只能难住人类。 如果公民真找到了网站上有用的部分，她还会遇到种种障碍，比如死链，或是印度铁路那令人抓狂的16字符姓名长度限制。 后台也好不到哪里去。 五月份，一名19岁的“道德黑客”暴露了一个重要的高中离校考试系统中的关键漏洞，引发了全国性的哗然。

On to your second question. It is indeed 2026 and the visa website of a country renowned for its people's IT skills resembles a Geocities page from 1999. One reason is that India's approach has long been to reproduce paper processes in digital form rather than rethinking them altogether. The concept of UX—jargon coined in 1993 to mean designing things to ease the user's experience—did not enter the official guidelines for government websites until 2023. Another reason is that the National Informatics Centre, the government's in-house technology provider (whose triumphs include the visa website), has little ability to push back on demands from ministries to fill their sites with everything except usable features.

接下来是你的第二个问题。 现在的确已经是2026年，而一个以其国民IT技能闻名的国家的签证网站，看起来却像是1999年的Geocities页面。 原因之一是，印度的做法长期以来一直是在数字形式下复制纸质流程，而不是从头重新思考整个流程。 UX（用户体验）这个概念——这个1993年创造的术语，意为通过设计来简化用户使用体验——直到2023年才被纳入政府网站的官方指南。 另一个原因是，国家信息中心——政府的内部技术提供方（其“杰作”就包括这个签证网站）——几乎无力抵制各部委提出的、要在网站上塞满除实用功能之外的一切内容的要求。

For projects beyond NIC's capabilities, officials look to the private sector. A mid-level bureaucrat is usually put in charge of procurement. Most have no technical knowledge. Their solution is to hire the best the market has to offer. They rely on brand-name consultants to help design the project and draft requirements. The most ambitious jobs are often executed by the same IT services companies that build products for Western clients.

对于超出国家信息中心能力的项目，官员们会向私营部门寻求帮助。 通常由一名中级官僚负责采购。 他们大多没有技术背景。 他们的解决方案就是聘请市场上最好的供应商。 他们依靠知名品牌咨询公司来帮助设计项目和起草需求。 最雄心勃勃的项目通常由同样为西方客户开发产品的IT服务公司来执行。

Yet here, too, the result can be lacklustre. By expecting outsourcers to do both the thinking and the execution on its behalf, the government puts itself in a vulnerable position, argues Susan Thomas of XKDR, a think-tank in Mumbai, in a recent paper. It buys a system "but lacks the internal expertise to understand what it has bought or how to evolve it, turning a strategic asset into a costly, unmanageable liability". Bureaucrats do not need AI to fall victim to cognitive surrender.

然而，即便在此，结果也可能乏善可陈。 孟买智库XKDR的苏珊·托马斯在最近的一篇论文中指出，政府指望外包商既替自己思考又替自己执行，这让自己陷入了弱势地位。 它购买了一个系统，“但缺乏内部专业知识来理解自己买的是什么，也不知道该如何让它演进，从而把一个战略资产变成了昂贵且难以管理的负担”。 官僚们不需要借助人工智能，就已经陷入了“认知投降”的境地。

As for your last question, official sites are unlikely to improve. They may soon offer better UX. But the real problem is institutional design rather than web design. Bureaucracy runs on risk aversion. Officials worried about anti-corruption bodies award contracts to the lowest bidder, despite having the discretion to weigh up other factors. They splurge on top-tier consultants because it gives them cover for their decisions. A bureaucrat who takes a risk and produces a terrific outcome gets no reward. One who does a bad job the correct way pays no penalty. Even the official who botched the school-leaving exams was merely transferred to the agriculture ministry.

至于你的最后一个问题，官方网站不太可能有所改善。 它们或许很快会提供更好的用户体验。 但真正的问题在于制度设计，而不是网页设计。 官僚体系的运作方式是规避风险。 担忧反腐机构的官员们会把合同授予出价最低的投标者，尽管他们本可以酌情权衡其他因素。 他们挥霍重金聘请顶级咨询公司，因为这能为他们的决策提供保护。 一个敢于冒险并取得出色成果的官僚得不到任何奖励。 一个按“正确”方式把事情搞砸的人也不会受到任何惩罚。 就连搞砸了高中离校考试的那位官员，也不过是被调到了农业部。

India is justly proud of itself for developing Aadhaar, a national biometric ID system that has become the foundation for many other digital services both public and private. But a huge reason for Aadhaar's success is that it was spearheaded by someone with both technical expertise and authority. Nandan Nilekani, a co-founder of Infosys, an IT-services firm, accepted the job on the condition that he had the formal status of a cabinet minister. That let him take calculated risks, think creatively and hire a bunch of Silicon Valley hotshots even as he stayed within the rules.

印度为自己开发的Aadhaar系统感到自豪，这是理所当然的。这个国家生物识别身份系统已成为许多其他公私数字服务的基础。 但Aadhaar成功的一个巨大原因在于，它由一位兼具技术专长和权威的人士牵头。 IT服务公司Infosys的联合创始人南丹·尼勒卡尼接受了这项工作，条件是他要拥有内阁部长的正式地位。 这让他得以在遵守规则的同时，承担经过计算的风险，进行创造性思考，并聘请了一批硅谷的顶尖人才。

Few governments can afford to pay their best techies what the private sector can. They may, though, lure talent with prestige, power and patriotism, as India did with Aadhaar. Yet it has failed to apply the lessons of what made that project a success. It is not the rules around building IT that need overhauling but the culture under which they are applied. Of that there is no question.

很少有政府能像私营部门那样，给自己最优秀的技术人才开出高薪。 不过，它们或许可以用声望、权力和爱国情怀来吸引人才，就像印度在Aadhaar项目上所做的那样。 然而，印度却未能将该项目成功的经验教训付诸应用。 需要彻底改革的，不是围绕建设IT系统的那些规则，而是应用这些规则时的文化。 这一点，是毫无疑问的。