永远不要忘记_它永远不会忘记一张脸

Most people have heard of facial recognition technology and probably have some concerns about the impact on privacy. But if you’re thinking, hey, the world’s a big place, or it’s easy to get lost in a crowd, or no one will be interested in me, let me stop you right there.

大多数人都听说过面部识别技术，并且可能会对隐私产生影响。 但是，如果您在想，嘿，世界是一个广阔的地方，或者很容易在人群中迷路，或者没人会对我感兴趣，那么让我在这停下来。

Someone does want to track you, and you cannot hide.

有人确实想追踪您，但您无法隐藏。

First, let’s understand just how effective facial recognition technology is today.

首先，让我们了解一下如今面部识别技术的有效性。

Facing facts

面对事实

Norway is the world’s biggest producer of farmed salmon. They raise more than 300 million fish every year and they all look exactly the same. Well, to us they do, but not to the artificially intelligent machine that is currently identifying the faces of individual salmon.

挪威是世界上最大的鲑鱼养殖国。 他们每年养殖超过3亿条鱼，而且看起来都一样。 好吧，对我们来说，它们确实如此，但对目前正在识别鲑鱼个体的人工智能机器却没有。

And it’s not just fish. Artificially intelligent vision can also recognize individual birds in the wild, along with chimpanzees, elephants, cows and pigs. Pets as well of course thanks to apps like Finding Rover.

而且不只是鱼。 人工智能视觉还可以识别野外的个体鸟类，黑猩猩，大象，牛和猪。 当然，宠物也要归功于诸如Finding Rover之类的应用程序。

So ask yourself, if facial recognition technology can identify the face of individual fish in the waters off the Norwegian coast, what chance do you have on a city street.

因此，请问问自己，如果面部识别技术可以识别挪威沿海水域中个别鱼类的面Kong，那么您在城市街道上有什么机会？

Starting to get worried? Here are some numbers to cheer you up.

开始担心？ 这里有一些数字可以使您振作起来。

• there will be 1 billion facial recognition cameras in the world by next year

  明年全球将有10亿台面部识别相机

• 1 in 2 Americans have been added to a law enforcement facial recognition network (and this was as of 2016)

  每2个美国人中就有1个被添加到执法人员面部识别网络中(截至2016年)

• police officers in one Florida county were running 8,000 facial recognition searches a month (and they didn’t need a reason to do so)

  佛罗里达州一个县的警员每月进行8,000次面部识别搜索(他们并不需要这样做)

Just the beginning

只是开始

And the US isn’t even close to being the most eager adopter of facial recognition. Let’s take a look at Singapore.

而且，美国甚至还不是最热衷于采用面部识别技术的国家。 让我们来看看新加坡。

The government is rolling out a system that already has more than four million Singaporeans inside its biometric database. While hotel guests are already checking in by looking at a camera and ID cards are expected to be replaced by 2022, they plan to install more than 100,000 sophisticated facial recognition cameras across the tiny country.

政府正在推出一个系统，该系统的生物识别数据库中已经有超过四百万新加坡人。 尽管酒店客人已经通过看摄像机来办理登机手续，并且身份证有望在2022年之前被取代，但他们计划在这个小国家安装超过100,000台复杂的面部识别摄像机。

Unsurprisingly, China is also enthusiastic about this technology. They have more than half of the world’s facial recognition cameras and a database that includes just about everyone in the nation. A leak from a database last year revealed that millions of people were being monitored daily by cameras at hotels, parks, tourism spots and mosques.

毫不奇怪，中国也对这项技术充满热情。 他们拥有全球超过一半的面部识别相机，并拥有一个数据库，其中包含了美国几乎所有的人。 去年从数据库中泄漏的数据显示，每天都有数百万人被旅馆，公园，旅游景点和清真寺的摄像头监视。

People in China have generally been more willing than Americans to sacrifice their privacy for security. But there are signs that facial recognition has gone too far. Even in China there is a growing unease about facial tracking with a poll showing that 3 in 4 people want to stick with regular ID rather than switch to face recognition.

通常，中国人比美国人更愿意为安全而牺牲自己的隐私。 但是有迹象表明面部识别已经走得太远了。 甚至在中国，关于面部跟踪的不安情绪也与日俱增。一项民意调查显示，每4个人中有3个人希望坚持使用常规ID，而不是改用面部识别。

This pushback is worldwide.

这种推销是全球性的。

San Francisco was the first US city to ban facial recognition in 2019 and last month IBM decided to abandon their work on the technology.

旧金山是2019年第一个禁止面部识别的美国城市，上个月IBM决定放弃他们在这项技术上的工作。

“We believe now is the time to begin a national dialogue on whether and how facial recognition technology should be employed by domestic law enforcement agencies.” — IBM CEO Arvind Krishna.

“我们认为现在是时候开始关于国内执法机构是否以及如何使用面部识别技术的全国性对话。” — IBM首席执行官Arvind Krishna 。

In the UK last week, a court ruled that police broke the law when they were scanning crowds at soccer matches to find people with warrants for their arrest — while also looking for people who were just vaguely of interest.

上周在英国，一家法院裁定，警察在足球比赛中扫视人群以寻找有逮捕证的人时，同时还寻找兴趣模糊的人，均违反了法律。

This fight over facial recognition is going to be a messy affair with a lot of competing interests. The EU, for example, had wanted a five-year ban on the technology, but then they withdrew the proposal earlier this year following objections from national security officials.

这场关于面部识别的斗争将与许多竞争利益混为一谈。 例如，欧盟曾希望对该技术实行五年禁令，但随后由于国家安全官员的反对，他们于今年早些时候撤回了这项提议。

Not all bad

并非都不好

And to be fair, facial recognition does have its uses that go beyond identifying a fish and finding rover. Criminals have few places to hide, lost children can be located quickly and as Singapore is finding out, it makes transactions involving money and ID so much easier and safer.

公平地说，人脸识别的作用不仅仅在于识别鱼类和寻找漫游者。 犯罪分子没有什么地方可以躲藏，失踪的儿童可以很快找到，新加坡正在发现，这使得涉及金钱和身份证的交易变得更加容易和安全。

So while some jurisdictions are saying no to facial recognition, over time it may be tough to stop as more and more places get hooked on the technology. Moscow, for example, is using its 100,000 cameras to enforce quarantines during the Covid-19 pandemic, while Sydney airport is testing it on passengers.

因此，尽管一些司法管辖区拒绝承认面部识别，但随着时间的流逝，随着越来越多的地方迷上该技术，可能很难停止。 例如，在Covid-19大流行期间，莫斯科正在使用其100,000台摄像机实施隔离，而悉尼机场正在对乘客进行测试。

In Tokyo, the Olympic Games may have been delayed till next year, but facial recognition tech will still be used to access events.

在东京，奥运会可能会推迟到明年，但面部识别技术仍将用于访问赛事。

And of course money talks. The facial recognition market was worth $3.2 billion last year and this will jump to $7 billion by 2024.

当然还有金钱的话题。 去年，面部识别市场价值 32亿美元，到2024年将跃升至70亿美元。

A fashionable response

时尚的回应

So the benefits of this technology may be too good to pass up. Just as we’ve grudgingly accepted location tracking in our phones and cookies on our computers, we may get used to facial recognition as the price you pay for progress.

因此，这项技术的好处可能太好了，难以超越。 就像我们在手机和计算机上的cookie中勉强接受了位置跟踪一样，我们可能已经习惯了面部识别作为您为进步付出的代价。

But this doesn’t mean you have to completely submit to the machine.

但这并不意味着您必须完全提交到计算机。

People are already coming up with makeup that can defeat some recognition software, while the AI startup Kneron discovered that a simple printed mask of someone else’s face was able to trick some systems.

人们已经想出了可以击败某些识别软件的化妆品 ，而AI初创公司Kneron发现， 别人脸部的简单印刷面具可以欺骗某些系统。

You can also buy sunglasses that claim to deceive the cameras, while others are selling T-shirts with images that distract the all-seeing eye from looking at your face.

您还可以购买声称会欺骗相机的太阳镜 ，而其他人则在出售 T恤衫，其图像会分散全视线，使您无法看着自己的脸。

So this is where we are today. We can beat facial recognition with simple hacks like masks — for now — but cameras are probably coming to most street corners.

所以这就是我们今天的位置。 目前，我们可以通过简单的黑客手段(例如面具)来击败人脸识别，但是相机可能会出现在大多数街角。

History shows us that people have never been able to resist new technology with so many clear benefits. There will be bitter fights ahead over how it’s used, but at the end of the day it’s probably just getting started.

历史向我们表明，人们从来没有能够抗拒具有如此众多明显好处的新技术。 关于如何使用它，将会有激烈的斗争，但是到最后，它可能才刚刚开始。