大数据作恶时代,比预期来得更快

科技 5个月前 雷科技
67 0 0
【腾讯云】多款云产品1折起,买云服务器送免费机器,最长免费续3个月

最近几年,大数据是一个相当热门的概念。在和AI等新潮技术结合后,它更是时常出现在各路互联网厂商的营销文案中。借助它,科技企业为我们描绘出一副技术改变世界、每位用户的个性化需求被无微不至满足地满足的美好世界。

然而,科技发展史告诉我们,技术通常都有两面性,在给人们带来福祉的同时,也可能会在某些方面产生负面效果。大数据也好,AI也罢,同样不能摆脱这个规律。而且,它们的负面作用,已经在影响我们的日常生活了。

给棋牌游戏玩家推送借贷广告

实际下,写这篇文章的契机,源于小雷最近的一次游戏体验。我们都知道,棋牌类游戏一直拥有不少的拥趸。而且,这类游戏开发难度小、性能要求低,还天然拥有大量潜在用户。小雷当时打了几把斗地主,几轮刺激的加倍、抢地主,一把梭哈后,平台赠送的金币很快见底。

对于想要获取更多金币的玩家,平台贴心地提供了两种方式。一种是直接氪金,购买金币,适合财大气粗或者输红了眼的玩家。另外一种也是熟悉的互联网模式——看广告换金币,这当然也无可厚非。只是,小雷仔细看了几轮广告后,惊讶地发现,大部分广告都是借贷类型的。广告的形势很多样,但主旨都是一个:借钱很容易、借钱好处多,快来找我借钱。

大数据作恶时代,比预期来得更快

(图源手机App截屏)

说实话,如果平时看到借贷类广告,小雷不会很感到奇怪。但如果在某个场景里总是精准推送借贷这一类广告,好巧不巧,这还是个棋牌游戏平台,就很容易让人怀疑是大数据的力量了。

对棋牌类游戏心存芥蒂,说白了还是因为它过去、现在乃至未来都是赌博这一违法活动的温床。其实,进入互联网时代,市面上不少来路不明的棋牌游戏,实质上都是网络赌博的载体。它们中伪装成金币的筹码,实际上是和真实货币形成隐射、可以相互转换的。

对于正规的棋牌类游戏平台,没有必要将它和赌博等同起来。只是,当它和精准的借贷广告连接起来时,所产生的“概率叠加”的现象,就难免让人浮想联翩。游戏平台和广告商至少有一方认为棋牌游戏玩家和借贷用户是高度重合的。

大数据不作恶?有点难

以上起到的大数据力量的案例,还是比较隐蔽,可能不是那么有说服力。但我们还能找到更多用户被大数据侵扰的痕迹,比如说大数据杀熟、算法沉迷、推荐信息茧房等。

大数据杀熟

大数据杀熟,大家应该不陌生了。早前有关部门出台了法律法规明文禁止互联网平台的杀熟行为,而各大互联网巨头也都承诺不会利用大数据进行杀熟。但遗憾的是,时至今日,大数据杀熟明面上可能消失了,但隐形的杀熟现象仍然非常普遍。

以小雷个人的经历来看,目前主要杀熟的方式是优惠券。一般来说,不同用户看到的商品价格是一致的,但能否能领取到优惠券、能领到多大金额的优惠券等,不同人的情况可能天差地别,最终为商品付费的价格也是不一样的。很显然,这也是一种大数据杀熟的方式。

大数据作恶时代,比预期来得更快

(图源新华社视频截图)

比如外卖平台,小雷身边很多朋友都是同时用几个外卖账号,哪个号优惠券金额大用哪个、哪个号会员续费便宜用哪个。

时间杀手

移动互联网带来了爆炸式的信息增长,互联网企业的重要工作之一就是把用户留在平台上。而互联网内容公司则需要用各种手段来让用户不断停留在内容上。

以现在很多的短视频平台来说,时长较短的视频没有进度条功能,时长稍长的视频的进度条颜色很浅。有的短视频平台,甚至取消了长视频App中常见的滑动调节音量、亮度快捷手势操作,上下滑动直接变成了切换到下一个视频。总之,这么多骚操作,就是为了让用户在视频平台上消磨、投入更多的时间。

信息茧房

移动互联网时代,不少人会感觉到算法推荐的内容越来越匹配自己的喜好。但长远来看,这并不是好事。它本质上是利用大数据识别出目标人群,然后根据年龄、性格、价值观等多重因素进行更加细化的分类,从而推送高匹配内容。久而久之,用户难以获取到全面的信息,以至于困在信息茧房中而不自知。

互联网出现和发展之初,很多人对它大加赞赏,认为互联网消除了物理空间的隔阂,让信息流动速度得到指数级的提升,进一步消灭信息差,从而消灭信息茧房。但现在来看,在商业利益的驱使下,互联网技术同样可以走向加厚信息茧房的方向。

隐私泄露

大数据的基础是海量的数据收集,这个过程中不可避免地要收集大量的个人用户信息。但如果这些数据保存或利用不当,产生的后果将是灾难性的。近几年,用户信息泄露事件层出不穷,比如剑桥分析公司通过Facebook窃取了8700万用户的私人信息,Google+平台存在的漏洞导致近5000多万个用户的信息面临泄露风险,国内一家大型酒店集团5亿条个人信息被放在暗网上售卖……

“在互联网上,没人知道你是一条狗”,这曾是远古互联网时代的一句名言。但说这句话的人,估计怎么也想不到未来的互联网时代,没有任何人的信息数据是安全的。

抵抗大数据作恶,用户能做的不多

大数据本质上只是一种技术,至于发挥正面还是负面效果,终究还是得看怎么用。而大数据作恶,大部分情况下都是商业利益考量战胜了用户权益保障。理论上说,用户拥有选择商家和服务方的权利。

但现实中,随着互联网服务的聚合乃至垄断,我们的衣食住行、工作生活等方面和互联网的连接,都掌握在少数互联网巨头手中。用户的选择权并不存在,用脚投票也就无从谈起。

中学阶段的经济常识就告诉我们,市场调节失灵时,就需要宏观干预。对付贪婪的巨头,最有效的手段是监管,让作恶的成本超过作恶的收益。Facebook的剑桥事件,最终以Facebook认罚50亿美元和解暂告一段落;Google为Google+相关诉讼支付了数千万,因cookie不当利用被开出1亿欧的罚款。

实际上,国内监管部门相应的手段也在不断增强。去年年底,《中华人民共和国个人信息保护法》正式施行;今年年初,《互联网信息服务算法推荐管理规定》公布,明确禁止杀熟、算法歧视等现象。从去年到现在,国内多个部门进入数家企业进行安全检查,调查其涉及的个人信息安全问题。

当达摩克利斯之剑持续悬在头顶时,相信大数据作恶会被有效遏制。

However, the history of the development of science and technology tells us that technology usually has two sides, which not only brings benefits to people, but also may have negative effects in some ways. Neither big data nor AI can get rid of this law. Moreover, their negative effects are already affecting our daily life.

Push loan advertisements to board games

In fact, the opportunity to write this article stems from Xiao Lei's recent game experience. As we all know, board games have always had a lot of fans. Moreover, this kind of game development difficulty is small, the performance requirement is low, but also naturally has a large number of potential users. At that time, Xiao Lei fought against the landlord, doubled and robbed the landlord in several rounds of excitement. After a handful of Soha, the gold coins given by the platform soon bottomed out.

For players who want to get more gold coins, the platform intimately provides two ways. One is direct krypton gold, the purchase of gold coins, suitable for deep-pocketed or red-eyed players. Another is also familiar with the Internet model-watching advertisements for gold coins, which is of course understandable.However, after watching several rounds of advertisements carefully, Xiao Lei was surprised to find that most of the advertisements were loan types. The situation of advertising is very diverse, but the theme is all the same: it is easy to borrow money, it is good to borrow money, come and borrow money from me quickly.

(screenshot of App of Picture Source Mobile phone)

To tell you the truth, Xiao Lei would not be very surprised if he saw loan advertisements. However, if advertisements such as loans are always accurately pushed in a certain scene, coincidentally, this is also a board game platform, it is easy to suspect that it is the power of big data.

To put it bluntly, it is because it was, is and will be a hotbed of illegal gambling. In fact, entering the Internet era, many unknown board games on the market are essentially the carriers of online gambling. The chips disguised as gold coins are actually implicit and interchangeable with the real currency.

For the formal board game platform, it is not necessary to equate it with gambling. However, when it is connected with accurate loan advertisements, the phenomenon of "probability superposition" will inevitably make people imagine.At least one side of the game platform and advertisers believe that board gamers and borrowers are highly overlapped.

Big data does not do evil? It's a little difficult.

The above case of big data's power is still relatively hidden and may not be so convincing. But we can also find more traces that users have been harassed by big data, such as big data's familiarity, algorithm obsession, recommended information cocoon room and so on.

Big data kills well.

Big data is familiar, and we should be no stranger to him. Earlier, relevant departments issued laws and regulations that explicitly prohibit the killing of cooked goods on Internet platforms, and the major Internet giants have also promised not to use big data to kill cooked goods. However, it is a pity that up to now, big data may have disappeared, but the invisible phenomenon is still very common.

According to Xiao Lei's personal experience, the main way to kill well at present is coupons. Generally speaking, the prices of goods seen by different users are the same, but whether they can get coupons, how much coupons they can get, and so on, the situation of different people may be very different, and the final price of paying for the goods is also different. Obviously, this is also a way for big data to kill well.

(video screenshot of Picture Source Xinhua News Agency)

For example, on the takeout platform, many friends around Xiao Lei use several takeout accounts at the same time, which number has a large amount of coupons, and which number is cheaper to renew.

Time killer.

Mobile Internet has brought explosive information growth, and one of the important tasks of Internet enterprises is to keep users on the platform. On the other hand, Internet content companies need to use a variety of means to keep users staying on the content.

For many of today's short video platforms, short videos have no progress bar function, while longer videos have a light progress bar. Some short video platforms even cancel the common sliding volume adjustment and brightness quick gestures in long video App, swiping up and down to switch to the next video.In short, so many coquettish operations, is to let users spend more time on the video platform.

Information cocoon room

In the era of mobile Internet, many people will feel that the content recommended by the algorithm more and more matches their preferences. But in the long run, this is not a good thing. In essence, it uses big data to identify the target population, and then makes a more detailed classification according to multiple factors such as age, personality, values and so on, so as to push the high matching content. Over time, it is difficult for users to get comprehensive information, so that they are trapped in the information cocoon room without knowing it.

At the beginning of the emergence and development of the Internet, many people praised it, thinking that the Internet has eliminated the estrangement of physical space, increased the speed of information flow exponentially, and further eliminated the information gap, thus eliminating the information cocoon house.But now, driven by commercial interests, Internet technology can also move towards the direction of thickening the information cocoon house.

Privacy disclosure

Big data is based on massive data collection, and it is inevitable to collect a large amount of individual user information in this process. However, if these data are not preserved or used properly, the consequences will be catastrophic.In recent years, user information disclosure incidents have emerged one after another. For example, Cambridge Analytics stole 87 million users' private information through Facebook, loopholes in the Google+ platform put nearly 50 million users at risk of disclosure, and 500 million pieces of personal information of a large domestic hotel group were sold on the secret Internet.

"on the Internet, no one knows that you are a dog", which was a famous saying in the ancient Internet era. But the person who said this sentence, it is estimated that no one's information and data is secure in the future Internet era.

There is not much users can do to resist big data's evil.

Big data is essentially just a technology, as to play a positive or negative effect, after all, it depends on how to use it. And big data did evil, in most cases, commercial interests considerations prevailed over the protection of users' rights and interests. In theory, users have the right to choose merchants and service parties.

But in reality, with the convergence and even monopoly of Internet services, the connection between our clothing, food, housing, transportation, work and life and the Internet is in the hands of a small number of Internet giants. Users do not have the right to choose, so it is impossible to vote with their feet.

Economic common sense in middle school tells us that when market regulation fails, macro intervention is needed. The most effective way to deal with greedy giants is regulation, so that the costs of doing evil outweigh the benefits of doing it. The Cambridge case of Facebook ended with a $5 billion fine from Facebook; Google paid tens of millions of dollars for Google+-related lawsuits and was fined 100m for improper use of cookie.

In fact, the corresponding means of domestic regulatory authorities are also increasing. At the end of last year, the "personal Information Protection Law of the people's Republic of China" was formally implemented, and at the beginning of this year, the "regulations on the recommendation and Management of Internet Information Service algorithms" was promulgated, explicitly prohibiting familiarity, algorithmic discrimination and other phenomena. From last year to the present, a number of domestic departments have entered several enterprises to conduct security inspections to investigate the personal information security issues involved.

When the sword of Damocles continues to hang over his head, I believe that big data's evil will be effectively curbed.

【腾讯云】多款云产品1折起,买云服务器送免费机器,最长免费续3个月
版权声明:雷科技 发表于 2022-07-03 14:06。
转载请注明:大数据作恶时代,比预期来得更快 | ColaBug

相关文章

colabug网址导航在线提交网站

暂无评论

-游客点评-
等待评论...