A new way to work
What will business technology look like tomorrow?
The Economist(July 13th. 2017)-Print Edition-Books and arts
Machine, Platform, Crowd: Harnessing our Digital Future.By Andrew McAfee and Erik Brynjolfsson.W.W. Norton; 402 pages; $28.95 and £22.99.
2014年，来自MIT的Andrew MaAfee与Eric Brynjolfsson联合发表了一本名为《第二次机器革命》的书。在这本书里作者理智客观地描述了新数码技术如何平衡推进社会发展，即便在这过程里失业率有所升高、工资水平有所下降。在他们的最新作品《机器，平台，大众》中，作者试图解释这些因素背后的商业图景。
Mr McAfee与Mr Brynjolfsson相信计算机与互联网的最新阶段给工作模式创造了三个变革。首先是人工智能（AI）：从人到机器的变革。在过去，人利用计算机工作的同时也被计算机提高，作者称之为“标准合作关系”。但是随着计算机的发展这个模式逐渐瓦解，计算机取得了更多控制权。
你只需要想一下自动驾驶汽车、在线语言翻译和亚马逊的无人值守商店原型[注1]就知道大的变化已经不远了。数码技术一度应用于信息领域，数字和文本的应用在先，带领音乐以及视频服务在后。现在数码技术已经进入到实体世界。
比如，设计一个用在冰箱等设备上的“热交换器”要考虑平衡诸多的技术参数以及限制，要优化好参数如此之难，大家也就做好了“何必完美”的打算。但是新“创成式设计”意味着集成AI的软件可以通过运行海量设计方案以寻找可能的最优设计——人可能永远没法做到这一点。利用3D打印，这些设计可以在任何地方被共享、修改和生产。
第二个改变是从产品到平台的转变。这样的例子每个人每天都遇得见。Uber：最大的出租车服务自己没有车；AirBnb：最大的旅馆服务自己没有房产；Alibaba：最完整的零售业自己没有货物；Facebook：最具价值的媒体公司自己却没有创造太多媒体内容；App Store：拥有220万个app，有苹果自己开发的App却少之又少。
企业通过构建平台去创造市场，从而让交易双方都能繁荣发展，像园丁一样的企业从稳定的收入来源中得到满足。可这些并不容易。平台必须保证高标准，要能吸引到不同类型的多个参与者（就像租车服务，平台一端是司机或App开发方，另一端是乘客或其他参与者）。但平台一旦运行起来，因为数码形式下可观的尺度，价值不可小觑。这本书最重头的戏码在于，利用大量的需求曲线图表，对平台经济的分析。
第三个改变是从“点”到“面”的转变。“点”代表着集中化的机构，比如中央银行，抑或“大英百科全书”；“面”则意味着去中心化，参与者自行组织，如同比特币管理虚拟货币的节点，或是Wikipedia的撰写人。
当交易成本较高时，就像Ronald Coase曾说的，公司决策会偏向对内[注2]。可数码技术降低了相互作用成本，很多事情可以通过非正式的组织实现。这带来了更大的尝试和创新。“核心组织通常会错配其所面对的挑战和机遇，而群体因为更大的体量，几乎不会产生这样的问题”，作者如是说。
对于这本书，可能学究们会说，这书只是基于单独话题，而且其他人对此认识深入的多。有些读者则会惊诧于章节末的要点总结和问题，总是揭露商业经典的虚假之处。但是请容忍这些。想要驾驭重要的数码趋势，《机器，平台，大众》这本书当属不二选择。
注1：在中国已经实现了这样的便利店，初见苗头；
注2：此处所指代内容可能偏向于业务或资产等安排。科思定理阐述交易费用及其与产权安排的关系，提出了交易费用对制度安排的影响。译者非经济专业，见谅。
原文：
IN 2014 Andrew McAfee and Erik Brynjolfsson of the Massachusetts Institute of Technology published “The Second Machine Age”. The book was a balanced portrait of how new digital technologies were poised to improve society, even as they increased unemployment and depressed wages. In their latest work, “Machine, Platform, Crowd”, the authors seek to explain the business implications behind these developments.
Mr McAfee and Mr Brynjolfsson believe that the latest phase of computers and the internet have created three shifts in how work happens. The first is artificial intelligence (AI): a move from man to machine. In the past people worked with computers and, at the same time, were augmented by them: what the authors call the “standard partnership”. But that model is breaking down as computers improve and take more control.
You need only look at self-driving cars, online language translation and Amazon’s prototype cashierless shops to see that something big is happening. Digital technologies used to be applied to information—first numbers and text, and, later, music and video. Now, the digital technologies are invading the physical world.
For instance, designing a “heat exchanger”, a part in appliances like refrigerators, means balancing many different specifications and constraints. Humans settle for one that works well enough because to find the optimal one is too hard. But new “generative design” means AI-infused software can run zillions of tiny permutations to find the best possible design—one that a human might not come up with. And with 3D printing, those designs might be shared, modified and manufactured anywhere.
The second is a shift from products to platforms. Many people encounter evidence of this every day. The largest cab service owns no vehicles (Uber), the biggest hotelier has no property (Airbnb), the most comprehensive retailer holds no inventory (Alibaba) and the most valuable “media” company creates some content but not much (Facebook). There are more than 2.2m apps in Apple’s store, almost none of which the company developed itself.
Platforms are a way for companies to create marketplaces that allow both sides of the transaction to flourish—while the firm, as gatekeeper, enjoys a tidy revenue stream. This is hard to pull off. The platform must ensure that standards are high, and also attract different sets of participants (like drivers or app developers on one side and customers on the other). But platforms are very valuable when they work, since they scale beautifully in a digital setting. The meatiest part of the book is the treatment of platform economics, replete with demand-curve charts.
The third shift is from the core to the crowd. The core refers to centralised institutions, like central banks or the “Encyclopedia Britannica”; the crowd refers to the decentralised, self-organising participants, be it bitcoin nodes that manage the virtual currency or contributors to Wikipedia.
When transaction costs are high, companies do things internally, as Ronald Coase, an economist, once noted. Yet as digital technologies lower the cost of interacting, more things can be done by informal groups. This leads to greater experimentation and innovation. “The core is often mismatched for the kinds of challenges and opportunities it faces, while the crowd, because it’s so big, almost never is,” the authors write.
Pedants will quibble that the book is built on individual themes that others have looked at more deeply. Some readers will be aghast that chapters end with bullet-point summaries and questions, evoking the worst of unctuous business tomes. But tolerate this. For an astute romp through important digital trends, “Machine, Platform, Crowd” is hard to beat.
