听力课堂TED音频栏目主要包括TED演讲的音频MP3及中英双语文稿，供各位英语爱好者学习使用。本文主要内容为演讲MP3+双语文稿:人工智能发展潜力太过可怕，未来是福是祸?，希望你会喜欢！

【演讲人】Leila Pirhaji

【演讲主题】《人工智能和代谢产物的潜力》

【演讲文稿-中英文】

翻译者 Carol Wang 校对 Jiasi Hao

In 2003, when we sequenced the human genome, we thought we would have the answer to treat many diseases. But the reality is far from that, because in addition to our genes, our environment and lifestyle could have a significant role in developing many major diseases.

在 2003 年， 当我们测序人类基因组时， 我们以为会找到 治疗多种疾病的答案。 但是实际情况远非如此， 因为除了我们的基因， 我们的生存环境和生活方式 也可能导致多种重大疾病。

One example is fatty liver disease, which is affecting over 20 percent of the population globally, and it has no treatment and leads to liver cancer or liver failure. So sequencing DNA alone doesn't give us enough information to find effective therapeutics.

例如，影响全球 超过 20％ 人口的脂肪肝， 没有任何有效的治疗方法， 而且最终可发展为肝癌 或肝衰竭。 因此，单纯的 DNA 测序 无法提供足够信息， 帮助我们寻找有效治疗方法。

On the bright side, there are many other molecules in our body. In fact, there are over 100,000 metabolites. Metabolites are any molecule that is supersmall in their size. Known examples are glucose, fructose, fats, cholesterol -- things we hear all the time. Metabolites are involved in our metabolism. They are also downstream of DNA, so they carry information from both our genes as well as lifestyle. Understanding metabolites is essential to find treatments for many diseases.

好消息是，我们体内 还有许多其他分子。 实际上，有超过 10 万多种代谢物。 代谢物是体积很小的分子， 像我们常听说的： 葡萄糖、果糖、脂肪、胆固醇等。 代谢物参与我们的新陈代谢， 它们在 DNA 的下游， 因此，它们携带着来自基因 和我们生活方式的信息。 了解代谢物对寻找许多疾病的 治疗方法至关重要。

I've always wanted to treat patients. Despite that, 15 years ago, I left medical school, as I missed mathematics. Soon after, I found the coolest thing: I can use mathematics to study medicine. Since then, I've been developing algorithms to analyze biological data. So, it sounded easy: let's collect data from all the metabolites in our body, develop mathematical models to describe how they are changed in a disease and intervene in those changes to treat them.

我一直想治病救人， 但尽管如此，在十五年前， 我因为喜欢数学 而离开了医学院。 不久之后，我发现了最酷的东西： 我可以使用数学来研究医学。 从那时起，我一直在开发 用于分析生物学数据的算法。 这听起来很简单： 我们先收集体内所有代谢物， 然后，开发数学模型 描述疾病中的代谢物变化， 并通过干预这些变化来进行治疗。

Then I realized why no one has done this before: it's extremely difficult.

然后，我终于明白 以前为何没人做这件事了： 这真是太困难了。

（Laughter）

（笑声）

There are many metabolites in our body. Each one is different from the other one. For some metabolites, we can measure their molecular mass using mass spectrometry instruments. But because there could be, like, 10 molecules with the exact same mass, we don't know exactly what they are, and if you want to clearly identify all of them, you have to do more experiments, which could take decades and of dollars.

我们体内有许多代谢产物， 种类繁多。 对于某些代谢物，我们可以使用 质谱仪来检测其分子量。 而质量完全相同的分子 可能有 10 种之多， 我们分不清谁是谁， 如果想识别所有这些分子， 则必须进行更多实验， 这可能需要几十年、 数十亿美元。

So we developed an artificial intelligence, or AI, platform, to do that. We leveraged the growth of biological data and built a database of any existing information about metabolites and their interactions with other molecules. We combined all this data as a meganetwork. Then, from tissues or blood of patients, we measure masses of metabolites and find the masses that are changed in a disease. But, as I mentioned earlier, we don't know exactly what they are. A molecular mass of 180 could be either the glucose, galactose or fructose. They all have the exact same mass but different functions in our body. Our AI algorithm considered all these ambiguities. It then mined that meganetwork to find how those metabolic masses are connected to each other that result in disease. And because of the way they are connected, then we are able to infer what each metabolite mass is, like that 180 could be glucose here, and, more importantly, to discover how changes in glucose and other metabolites lead to a disease. This novel understanding of disease mechanisms then enable us to discover effective therapeutics to target that.

为了做这件事，我们开发了 人工智能（AI）平台， 我们利用生物数据的增长， 建立了一个数据库， 包含代谢物现有信息 及与其它分子的相互作用的数据。 我们将所有这些数据 组合成了一个大型网络， 然后，在患者的组织或血液中， 测量代谢物的质量， 并寻找因疾病 而产生变化的代谢物的质量。 但是，正如我之前提到的， 我们并不知道是什么代谢物。 分子量为 180 的代谢物 可以是葡萄糖、半乳糖或果糖， 在我们体内，它们的质量完全相同， 但功能不同。 我们的 AI 算法 考虑了所有这些可能。 然后，会挖掘那个巨型网络的数据， 以发现那些代谢物 如何相互关联而导致疾病。 根据它们的关联方式， 我们就能推断出 每个代谢物的质量， 如 180 分子量的 可能是葡萄糖， 更重要的是， 发现葡萄糖和其他代谢物的变化 如何导致疾病。 对疾病机制的这种新颖理解， 使我们能够发现 针对该疾病的有效疗法。

So we formed a start-up company to bring this technology to the market and impact people's lives. Now my team and I at ReviveMed are working to discover therapeutics for major diseases that metabolites are key drivers for, like fatty liver disease, because it is caused by accumulation of fats, which are types of metabolites in the liver. As I mentioned earlier, it's a huge epidemic with no treatment.

凭借该技术，我们成立了 一家初创公司， 将该技术推向市场， 进而影响人们的生活。 现在，我们的 ReviveMed 团队 正努力寻找主要代谢疾病的疗法， 例如脂肪肝， 因为它由脂肪堆积造成， 而脂肪是肝脏中的代谢物。 如前所述，这种大型流行病 尚无有效疗法。

And fatty liver disease is just one example. Moving forward, we are going to tackle hundreds of other diseases with no treatment. And by collecting more and more data about metabolites and understanding how changes in metabolites leads to developing diseases, our algorithms will get smarter and smarter to discover the right therapeutics for the right patients. And we will get closer to reach our vision of saving lives with every line of code.

脂肪肝只是其中一个例子， 展望未来，我们将研究其它几百种 尚无有效疗法的疾病。 通过收集更多代谢物的数据， 了解代谢物的变化 如何导致疾病发展， 我们的算法会逐步完善， 为某些患者找到合适的疗法。 而且，我们将更加接近我们的愿景： 用程序代码拯救生命。

Thank you.

谢谢。

（Applause）

（掌声）