科技巨头联手，制定开放式医疗标准

不久前，多家顶尖科技公司商定按统一标准打造医疗保健产品，方便医院更共享医疗数据。开放式标准和规格的另一目标是鼓励“无障碍数据交换”。

对医疗专业人员和患者来说，亚马逊、谷歌、IBM、微软、甲骨文和Salesforce达成的这项协议是个好消息，因为掌握更多信息可能有助于迅速实施诊治。如果设想中的高效率成为现实，或许还能节省不少成本。

就医疗保健IT市场今后的竞争力而言，这些标准的开放性也是非常好的消息，因为对规模不如当前这些企业的公司来说，市场准入门槛会降低。

另外，这项协议显然有利于签约各方，不仅因为此举将扩大潜在市场，而且有助于实现人工智能方面的远大目标。如果处理得当，更多数据就能让人工智能更聪明。

各家公司确实都需要更聪明的人工智能。想想IBM和沃森大系统，沃森最近还成了《华尔街日报》的抨击对象，主要内容是十几名客户以及合作伙伴已缩小合作规模，或终止用沃森治疗癌症的试验。

《华尔街日报》的报道称：“在许多情况下，人工智能工具并未创造出多大的价值。处理某些病历时沃森不够准确，会因为罕见或癌症复发缺乏数据而出错，治疗情况变化的速度明显超过人类培训师更新沃森系统的速度。”

这里的教训告诉我们要实际地看待AI目前的发展水平。但也可以做出另一种解读，也许对一家公司来说医疗保健领域过于复杂，无论多么令人尊敬，也不管资源多丰富，企业很难单打独斗提供所有价值。如果情况确实如此，就很有必要制定协议破除条块分割开放市场，而且很有希望挽救更多生命。（财富中文网）

Some of the biggest tech giants yesterday agreed to build health care products that use a common set of standards in order to make it easier to share medical data across hospitals. What’s more, these are open standards and specifications aimed at encouraging “frictionless data exchange.”

The agreement between Amazon, Google, IBM, Microsoft, Oracle and Salesforce is good news for medical professionals and their patients, because having more information at hand potentially aids speedy diagnosis and treatment. If the promised efficiencies materialize, there may even be savings on the horizon.

The openness of the standards is also very good news for the future competitiveness of the health care IT market, because it should lower the barriers to entry for companies that aren’t at the scale that these giants currently enjoy.

But the deal is also clearly good for its signatories, not only because it expands their potential markets, but because it will help feed their artificial intelligence ambitions. More data, if properly wrangled, translates to smarter AI.

And the companies need smarter AI. Consider the case of IBM and its marquee Watson system, which was recently the subject of quite the takedown by the Wall Street Journal. It turns out that more than a dozen clients and partners have scaled down or ended their experiments with using Watson in cancer treatment.

Here’s the relevant bit from that piece: “In many cases, the tools didn’t add much value. In some cases, Watson wasn’t accurate. Watson can be tripped up by a lack of data in rare or recurring cancers, and treatments are evolving faster than Watson’s human trainers can update the system.”

There’s a lesson in here about taking a realistic view of where AI’s development stands today. But there’s another takeaway too: perhaps health care is too sprawling an environment for one company, no matter how venerable and well-resourced, to provide all the value that’s needed on its own. If that’s the case, then there’s much cause to celebrate an agreement that breaks down silos, opens up markets, and hopefully saves lives.