Alphabet Inc’s Google routinely fields product queries from millions of shoppers. Now it wants to take a cut of their purchases, too.
The U.S. technology company is teaming up with retailers including Target Corp, Walmart Inc, Home Depot Inc, Costco Wholesale Corp and Ulta Beauty Inc.
Under a new program, retailers can list their products on Google Search, as well as on the Google Express shopping service, and Google Assistant on mobile phones and voice devices.
Bloomberg Markets; Suzi Ring , Gavin Finch , and Franz Wild
from
Christian Bittar was once among Deutsche Bank AG’s highest-paid traders, a math whiz who earned a near 90 million-pound ($126 million) bonus in 2008 alone. Now he’s sitting in a U.K. prison.
The 46-year-old former star banker pleaded guilty in a London court on March 2 to conspiring to rig the interest-rate benchmark known as Euribor. He’s in custody and will be sentenced after a related trial ends this summer. A court lifted reporting restrictions on his plea Thursday.
It’s a seminal moment for U.K. prosecutors in their six-year investigation. Bittar, famous for receiving multi-million dollar bonuses, is one of the highest-profile traders to be convicted in the global rate-rigging probe.
Artificial intelligence (AI) is going to impact several industries in a big way. Retailers, banks, carmakers, or technology companies, are scurrying to embrace AI to make their customers’ lives easier. Not surprisingly, the market for AI software is predicted to jump from just $3.2 billion a couple of years ago to $89.8 billion by 2025.
Tech giants like Baidu (NASDAQ: BIDU) have been pouring a lot of money into AI research. Let’s look at how Baidu plans to take advantage of the AI opportunity and why it could be one of the best bets in this space.
For more than a decade, the Chinese government has been working to push the Chinese manufacturing sector up the value chain. More recently, the push from the central government has become more formalized, resulting in the 2015 issuance of the State Council manufacturing modernization manifesto: Made in China 2025《中国制造2025》(State Council, July 7, 2015). Made in China 2025 focuses less on the types of products to be manufactured and more on the methods of manufacturing. It is okay to continue making rubber duckies, so long as the process for doing so is modernized. That is, massive automated factories churning out thousands of identical items with minimal human intervention.
The Chinese government has made clear it believes the largest and most successful manufacturing companies in the world have achieved that status in large part through software/information technology. This has led China to focus on artificial intelligence (人工智能). The Chinese government experienced what Will Knight at the MIT Technology Review has termed China’s AI Enlightenment. The process started with the issuance by the State Council of A Next Generation Artificial Intelligence Development Plan (新一代人工智能发展规划 July 8, 2017) setting forth a plan for AI development in China. The plan will progress in three stages, concluding in 2030. The final goal is ambitious: by 2030, China’s AI theories, technologies, and applications will lead the world, making China the world’s primary AI innovation center.
We are now in Stage 1 of the AI Plan, covering the period from 2018 to 2020.
Engineers are on the cusp of on-device machine learning, as evidenced by the first NIPS workshop on the subject in late 2017, and the advent of new neural processors, such as Kirin 970 from Huawei and Snapdragon 845 from Qualcomm.
Thus far, progress in artificial intelligence has been fueled primarily by the availability of data and more computing power. Classical machine learning has mostly been built on a single central node (usually in a data center) with full access to a global dataset and a massive amount of storage and computing power. Currently, many deep learning algorithms reside in the cloud, enabled by popular toolkits such as Caffe and TensorFlow, as well as specialized hardware such as tensor processing units.
But this centralized approach won’t work for things and applications that require low latency, such as flying a drone, controlling a self-driving car, or sending instructions to a robotic surgeon. To perform these delicate tasks, and other activities experts can’t yet anticipate, future wireless systems will need to make even more decisions at the network edge (closer to devices), more quickly and more reliably, even when they lose connectivity.
… The existing [sleep disorder] diagnostic procedure is labor-intensive and expensive, making it inaccessible to many people. It is also cumbersome and unpleasant; sleep labs haven’t changed much since the 1970s, said Reuveny, “You sleep outside your house. You are wired to 10-20 different electrodes. People are watching you, monitoring you during the night.”
While home diagnostic tests are cheaper ($200 per night compared to $1,000-$5,000 in-lab) there is still a lot of friction in the process: waiting lists can be long, patients may damage the device or fail to return it to the clinic, or they may have difficulty assembling it at home.
“If you combine all things together, you understand that something doesn’t work properly in the way people diagnose and manage sleep disorders today and this is where we come into play,” said Reuveny.
Credit Karma, the credit-checking startup that’s valued at over $3 billion, has acquired a young personal finance company, Penny, the companies tell Recode.
Penny is like other apps that allow users to track their cash and spending habits. But what makes the three-year-old, Y Combinator-backed startup different from better-known companies like Mint is its chat interface that allows the consumer to have a conversation with a bot and, for instance, ask questions about their month-over-month spending.
So Credit Karma is adding that interface to its arsenal as it seeks to move beyond merely offering credit scores and serve as a broader personal finance concierge. Now the company can combine its existing data on its customers with a user interface that allows Credit Karma to serve a bigger purpose as a personal finance coach.
How was a small app-based quiz used to harvest comprehensive data about that many people? At that point in history, Facebook’s API (the portal that allows third parties to make use of Facebook software and data) by default allowed third parties to access not only your own profile with permission, but also the full profiles of all of your friends.
Thus, a quiz app with 270,000 users could easily provide access to 50,000,000 full profiles. That represents only 185 friends per user, which is below average. According to the Times, 30 million of those profiles had enough information in them to correlate with other real-world datapoints held by data brokers and commonly used by political campaigns. This enabled Cambridge Analytica to connect these psychometric Facebook profiles to actual voters and offer their clients the ability to tailor advertisements to detailed psychometric profiles.
Facebook no longer allows such expansive access to friends’ profiles via the API and requires clearer explanations about what data APIs request access to.
So if the Facebook API allowed Kogan access to this data, what did he do wrong? This is where things get murky, but bear with us. It appears that Kogan deceitfully used his dual roles as a researcher and an entrepreneur to move data between an academic context and a commercial context, although the exact method of it is unclear.
Sen. Orrin Hatch March 15 continued his push for a bill that would give cloud computing companies, email service providers, and the U.S. government clarification on how law enforcement can access data stored abroad related to criminal probes of foreigners.
The CLOUD Act, ( H.R. 4943, S. 2383), introduced by Hatch (R-Utah) and Rep. Doug Collins (R-Ga.), also would allow the U.S. to enter into bilateral law enforcement data sharing agreements with foreign nations and clarify that requests for such data apply abroad.
The Institute for Advanced Study hosted a groundbreaking ceremony today for Rubenstein Commons, a new $20 million campus building that will provide a necessary space for enhanced communication and collaboration among Faculty and scholars at one of the world’s leading centers for curiosity-driven basic research. Rubenstein Commons, which is designed by Steven Holl Architects, is made possible by a gift from Institute Trustee David M. Rubenstein, renowned philanthropist and Co-Founder and Co-Executive Chairman of The Carlyle Group.
The groundbreaking ceremony was attended by David M. Rubenstein, architect Steven Holl, Princeton Mayor Liz Lempert, as well as Institute Trustees, Faculty, scholars, and staff.
Company Data Science News
Facebook is facing a Cook County lawsuit in the wake of the news that an app-maker running an app on its platform, Cambridge Analytica, had violated boundaries between academic and corporate research and obtained data from 50 million users. And only 207,000 of those users had plausibly given their consent. They consented to academic research, not corporate or political research, so it would be inaccurate to call their consent “informed.” I spoke with one of those 207,000 users who reported that he would never have taken that psychometric quiz except that he trusted Cambridge University, whose logo was used in the app. This user even took the time to match the name of the app maker with a professor at Cambridge to allay his misgivings. Other users are upset that when they downloaded the archive of all data Facebook holds about them, it included the calls they made from their Android phones, plus Messenger transcripts.
While I don’t see the following fact as a ‘gotcha’, it turns out that Aleksandr Kogan had co-authored papers with facebook research staff, which makes the companies attempt to place a majority of the blame on Kogan and Cambridge Analytica more complicated. This has led some researchers to call for Facebook to share more user data with academics (unlikely). Individuals are deleting Facebook. Investors are selling stock – it’s down ~15% from its month-long high. The Federal Trade Commission has launched an investigation. Facebook took out mea culpa advertisements in print newspapers around the US to continue their apology tour that started with Mark Zuckerberg’s uncomfortable interview on CNN last week. The Senate Judiciary Committee is joining the ranks of US and UK government bodies asking him to testify. DATA ETHICS MATTER.
Magenta is an open source machine learning music synthesis project that originated in the Google Brain team. One prototype, NSynth Super, is a collaboration with the Google Creative Lab. It uses, “a deep neural network to learn the characteristics of sounds, and then create a completely new sound based on these characteristics.” Google Creative Lab is now hiring.
Speaking of Google getting creative, they are trying to democratize surgery, a field I typically don’t think of when I think about the benefits of democracy. But, hey, I have been witness to a few kitchen sutures so bring it on, Google, bring on the democratic surgeries. (Full disclosure: my father is trained in surgical technique and the patient in our kitchen was a dog. She healed nicely and was the best, bravest dog ever. Twenty sutures, no anesthesia, no whimpering, no flinching.)
A fintech startup called Random Forest Capital has been acquired by Franklin Templeton. Random Forest touts its ability to mine “massive amounts of unstructured data” for insights and market predictions.
Bo Zhu, a research fellow in the Massachusetts General Hospital Martinos Centercreated AUTOMAP, an advancement in the medical imaging field that can use lower quality images and make decisions faster than existing image recognition methods. In some applications, the decisions can be made while the patient is still in the scanner, leading to a host of new opportunities in patient care.
Alphabet and Johnson & Johnson are partners in a start-up called Verb Surgical, a wildly ambitious project to develop robotics and machine learning tools to democratize surgery.
“The National Science Foundation’s Directorate for Social, Behavioral and Economic Sciences encourages submission of proposals that target reproducibility and replicability efforts in data-intensive domains and that specifically rely on analysis of neuroimaging or neuroelectric data, including but not limited to electroencephalography, magnetoencephalography, electrocorticography and functional neuroimaging.” Deadline for submissions is June 11.
“The $10 million ANA Avatar XPRIZE is a four-year competition focused on accelerating the integration of emerging and exponential technologies into a multipurpose avatar system that will seamlessly transport human skills and experience to the exact location where and when they are needed.” Registration closes on October 31.
I occasionally give free Go consults and code review on top of my daily work. As such, I tend to read a lot of other peoples’ codes. And while this is really more of a feeling *, I’ve seen an increase in what I call “Java-style” interface usage.
This blog post is a Go specific recommendation from me, based on my experiences writing Go code, on how to use interfaces well.
University of Washington, UW Institute on the Algorithmic Foundations of Data Science, John Thickstun
from
“The production of classical western music is a two stage process. First a composer writes down a score: written notation that indicates a particular musical structure. Then a performer reads this score and manipulates an instrument as indicated by the score to produce audio waves that a human ear perceives as music.” … “In this post, we will consider methods that replace the human transcriber in this loop with an automated algorithm.”
In this episode of the ARCHITECHT Show, internet of things expert Stacey Higginbotham joins me to discuss a wide range of topics regarding edge computing and the evolution of IoT infrastructure. Higginbotham discusses Cloudflare’s edge computing plans, the blurry nature of the actual edge, who stands to win and los from custom AI chips, and why 5G might not be the IoT savior it’s made out to be. [audio, 44:57]