AI Could Provide Moment-by-Moment Nursing for a Hospital’s Sickest Patients

Behnood Gholami, Wassim M. Haddad and James M. Bailey, AI Could Provide Moment-by-Moment Nursing for a Hospital’s Sickest Patients, IEEE Spectrum, 24 Sep 2018

At our company, Autonomous Healthcare, based in Hoboken, N.J., we’re designing and building some of the first AI systems for the ICU. These technologies are intended to provide vigilant and nuanced care, as if an expert were at the patient’s bedside every second, carefully calibrating treatment. Such systems could relieve the burden on the overtaxed staff in critical-care units. What’s more, if the technology helps patients get out of the ICU sooner, it could bring down the skyrocketing costs of health care. We’re focusing initially on hospitals in the United States, but our technology could be useful all around the world as populations age and the prevalence of chronic diseases grows. 

The benefits could be huge. In the United States, ICUs are among the most expensive components of the health care system. About 55,000 patients are cared for in an ICU every day, with the typical daily cost ranging from US $3,000 to $10,000. The cumulative cost is more than $80 billion per year.

Models Will Run the World

Steven A. Cohen and Matthew W. Granade, Models Will Run the World, Wall Street Journal, Aug. 19, 2018  

Tencent, the Chinese social-media giant and maker of WeChat , is one of our favorite examples of this new business model. A Tencent executive told us last fall: “We are the only company that has customer data across social media, payments, gaming, messaging, media, and music, and we have this information on [several hundred] million people. Our strategy is to put this data in the hands of several thousand data scientists, who can use it to make our products better and to better target advertising on our platform.” That unique data set powers a model factory that constantly improves user experience and increases profitability—attracting more users, further improving the models and profitability. That’s a model-driven business....

好好投資科技的 AI 深度學習演算系統

陳君毅，曾為上流社會理財，兩位外資金童拚創業，把頂級財富管理帶給普羅大眾，數位時代，2018.09.25

和碩布局車用電子

王郁倫，奧迪跨入電動車，和碩代工練兵十年、迎來春天，數位時代，2018.09.21

和碩技術長黃中于是和碩集團技術研發幕後推手，內部同仁暱稱「黃博」，他接受《數位時代》訪問指出，未來汽車將有四大趨勢，分別是：自駕車、新能源、通訊技術（如SOS智慧緊急求助Intelligent Emergency Call、道路救援Breakdown cover）及共乘。 

和碩布局車用電子多年，今年COMPUTEX上展示曲面玻璃接合技術，用在車內裝儀表板電腦，並預告開發自駕車關鍵行動車聯網技術V2X（Vehicle-to-everything）及5G車載通訊技術，該通訊技術由高通、奧迪、福特及5G汽車協會（5GAA）共同推出，是自駕車關鍵技術。 

V2X車聯網技術預計2020年成熟，被視為自駕車的關鍵技術，是因可以讓行駛於道路上的自駕車輛，對周遭的車輛、行人與各項設施進行資料傳輸與分享，讓駕駛提早獲得潛在危險警示，避免意外事故，比方在十字路口，預先知道對向車道有來車，特別的是，警訊靠的不是影像技術，而是通訊技術，所以不會有機器視覺「盲點」。...

Efficient tuning of online systems using Bayesian optimization by Facebook

Ben Letham, Brian Karrer, Guilherme Ottoni, and Eytan Bakshy, Efficient tuning of online systems using Bayesian optimization, Facebook Research, September 17, 2018

A/B tests are often used as one-shot experiments for improving a product. In our paper Constrained Bayesian Optimization with Noisy Experiments, now in press at the journal Bayesian Analysis, we describe how we use an AI technique called Bayesian optimization to adaptively design rounds of A/B tests based on the results of prior tests. Compared to a grid search or manual tuning, Bayesian optimization allows us to jointly tune more parameters with fewer experiments and find better values. We have used these techniques for dozens of parameter tuning experiments across a range of backend systems, and have found that it is especially effective at tuning machine learning systems.... 

We have used the approach described in the paper to optimize a number of systems at Facebook, and describe two such optimizations in the paper. The first was to optimize 6 parameters of one of Facebook’s ranking systems. These particular parameters were involved in the indexer, which aggregates content to be sent to the prediction models. The second example was to optimize 7 numeric compiler flags for HHVM. The goal of this optimization was to reduce CPU usage on the web servers, with a constraint on not increasing peak memory usage. 

解決教育機會不均等的契機在低年級及學前

曾世杰，解決教育機會不均等的契機，並不在高等教育階段，報導者，2018.9.12

早期介入，尤其是低年級及學前的介入，幼兒仍未有太多學習挫折的經驗，沒有習得無助的問題，也不會抗拒介入；即使真有落後，和優勢同儕比較起來，尚未落後太多。在執行技術層面，幼兒園的課程較有彈性，小一小二下午不必上課，介入時間容易安排，這些因素都可以降低介入的難度及成本，提高成功的機會。 

結論是，解決教育機會不均等的契機，並不在高等教育階段；升學制度改變，只能對現狀做出微調。真正會造成改變、真正能給弱勢學生帶來希望的，都在學前及國小教育階段。

北醫團隊用 AI 把關問題處方箋

蔡紀眉，「犯錯是人類的天性！」北醫團隊取經健保資料庫，用AI把關問題處方箋，數位時代，2018.04.20

用藥錯誤的情況有多嚴重？從數據來看，2013年美國因醫療疏失死亡的人數約為25萬人，高居全美前三大死因。但醫療疏失涵蓋甚廣，其中，因用藥不當導致死亡者，約在1萬7千人到3萬5千人之間。每年，美國政府需花費300億美元處理因用藥錯誤衍生的醫療問題。 

台北醫學大學醫學科技學院院長、萬芳醫院皮膚科主任李友專表示，原則上，用藥錯誤可能來自四個階段，依發生時序分別是處方錯、調劑錯、給藥錯以及病人錯誤服用。其中，最嚴重也最常發生的，是醫師在處方開立階段就出現錯誤，「這很難挽救，藥師也救不回來。」 

台灣方面，李友專預估，一年約有1千7百萬張的不適當處方，若以每人每年平均看診15次計算，台灣每年3億4千5百萬處方箋中，不適當處方率近5%，這個比例和美國相當，接近全球平均。...

雙連安養中心導入友達「智慧長照」

吳元熙，上千人搶住的爆紅安養院，為何需要導入友達「智慧長照」方案？，數位時代，2018.09.12  

這方案的一大亮點，在於能即時監測長輩的身體數據，上傳雲端同步至護理人員的平板電腦、手機與護理站，減輕傳統紙筆紀錄、半夜巡房的人力負擔。 

友達頤康智慧床墊，搭載壓力和振動兩種感應器，能隨時偵測長輩呼吸與心跳，也能設定警示範圍，提醒護理人員多加留意。特別的是，這款智慧床墊，能整合家中現有的床墊，不排除未來會單獨銷售，直接推向消費者市場。 

由於安養院裡有許多行動不便的老人家，友達也設計了室內定位按鈕，提供長輩隨身攜帶，系統則會在螢幕上顯示移動軌跡，定位誤差範圍小於30公分。若擔心長輩停留在同一位置過久，則能夠設定警示時間，讓系統發出推播警告。 

另外還有這台健康促進機，看起來跟一般健身房器材相似，卻能夠紀錄使用者的力道，逐漸變化阻力，最終達成穩定平衡，可以幫助長輩維持肌力；使用後，則會生成專屬的肌力報告，讓園區的復健治療師安排專屬運動課表。

Modeling the impact of AI on the world economy

Jacques Bughin, Jeongmin Seong, James Manyika, Michael Chui, and Raoul Joshi, Notes from the frontier: Modeling the impact of AI on the world economy, McKinsey Global Institute, September 2018.

Several barriers might hinder rapid adoption and absorption (see video, “A minute with the McKinsey Global Institute: Challenges of adopting automation technology”). For instance, late adopters might find it difficult to generate impact from AI, because front-runners have already captured AI opportunities and late adopters lag in developing capabilities and attracting talent. 

Nevertheless, at the global average level of adoption and absorption implied by our simulation, AI has the potential to deliver additional global economic activity of around $13 trillion by 2030, or about 16 percent higher cumulative GDP compared with today. This amounts to 1.2 percent additional GDP growth per year. If delivered, this impact would compare well with that of other general-purpose technologies through history.

最低的水果摘完之後

顏擇雅，最低的水果摘完之後，天下雜誌，2018

我寫這本書的初衷很單純，就是受不了大家在唱衰台灣。對我來說，台灣問題就是最低的水果摘完了，如今應該趕緊打造工具去摘更高的水果。

先進國家都老早摘完伸手能摘的水果，之後又經歷多次「摘完某一高度水果」的時刻。一九八二年我去美國念書，美國就處於如此窘境，媒體都是關廠、裁員新聞，大學生畢業即失業，街頭遊民暴增。但在此之前，自從一七七六年宣布獨立，美國已有過四十幾次衰退，每次都挺過來了，因此我沒聽到美國人自己在唱衰美國，或嫌年輕人一代不如一代。

人工智慧應用的兩大領域

羅耀宗，掌握人工智慧應用的兩大「錢途」，哈佛商業評論，數位版文章，2018/8/24

 Michael Chui, Nicolaus Henke, and Mehdi Miremadi, Most of AI's Business Uses Will Be in Two Areas, HBR, JULY 20, 2018 (from McKinsey (麥肯錫公司))

我們深入檢視19個產業和9個業務職能裡，超過四百個實際的人工智慧使用案例，發現有一句古老的格言，最能回答應將人工智慧用在何處的問題，那句格言就是：跟著錢走。 

傳統上，對企業提供最多價值的業務領域，往往是人工智慧能產生最大影響的領域。舉例來說，零售組織中，行銷和銷售經常提供重大的價值。我們的研究顯示，在顧客資料上使用人工智慧，將促銷活動個人化，單是實體零售商店的新增銷售（incremental sales）就會增加1％到2％。相較之下，在先進製造方面，營運活動往往產生最多的價值。這方面，人工智慧能夠協助根據需求背後的因果驅動因素，來做出預測，而不是根之前的結果來預測，因而改善預測準確度達10％到20％。這可能會使得存貨成本降低5％，營收提高2％到3％。