/* ---- Google Analytics Code Below */

Thursday, October 31, 2019

Being Paid for Your Data

More and better data means better training.   Better results.   And better adaption to future knowledge.   Still not perfect results, and the risk involved in that imperfection must be considered.  Payment or the means to use your data in other ways to improve your life.

AI Needs your Data and You Should get Paid for it.  by Gregory Barber in Wired

ROBERT CHANG, A Stanford ophthalmologist, normally stays busy prescribing drops and performing eye surgery. But a few years ago, he decided to jump on a hot new trend in his field: artificial intelligence. Doctors like Chang often rely on eye imaging to track the development of conditions like glaucoma. With enough scans, he reasoned, he might find patterns that could help him better interpret test results.

That is, if he could get his hands on enough data. Chang embarked on a journey that’s familiar to many medical researchers looking to dabble in machine learning. He started with his own patients, but that wasn’t nearly enough, since training AI algorithms can require thousands or even millions of data points. He filled out grants and appealed to collaborators at other universities. He went to donor registries, where people voluntarily bring their data for researchers to use. But pretty soon he hit a wall. The data he needed was tied up in complicated rules for sharing data. “I was basically begging for data,” Chang says.

Chang thinks he might soon have a workaround to the data problem: patients. He’s working with Dawn Song, a professor at the University of California-Berkeley, to create a secure way for patients to share their data with researchers. It relies on a cloud computing network from Oasis Labs, founded by Song, and is designed so that researchers never see the data, even when it’s used to train AI. To encourage patients to participate, they’ll get paid when their data is used.  ... "

More on McD's Knows What You Want

More on the McDonald's drive through idea.   Less friction, more sales.  Note plan to roll to all drive-throughs. Looking forward to see it in action.

McDonald’s drive-thru AI knows what you want before you orde by Matthew Stern in Retailwire with further expert comments.

Implementing in-store touch screen ordering kiosks was only the beginning for McDonald’s in efforts to modernize its operations with technology. One of the chain’s recently announced experiments involves using artificial intelligence (AI) to enable the drive-thru to recommend what customers might want before they place an order.

At some locations, McDonald’s has been testing a solution that recognizes the license plate of a customer in the drive-thru and uses that information to provide AI-based recommendations of menu items, as reported by The New York Times. The technology (which requires permission from customers) takes into account the customer’s previous order history when populating suggestions on the drive-thru touchscreen.

The solution adds another element of personalization to the already dynamic new drive-thru touchscreen at some McDonald’s locations, which tailor product recommendations based on big-picture factors like weather, wait time and item popularity. McDonald’s says that recommendation algorithms have already demonstrated an unspecified increase in order size and the chain plans to roll them out to all its drive-thrus in the U.S. by the end of the year.  ....  '

Predictions in Science

Could you crowdsource validity within science?

Individual predictions might not tell us much, but group predictions are useful.   in Arstechnica

Last year, a huge group of researchers collaborated to try to replicate the results of some very famous social science research. They determined that only 62% of the studies found similar results when they were repeated. But the researchers found something else intriguing: other scientists were astonishingly good at guessing which of the results would replicate.

Does that mean we can just ask scientists for their hunch on what research is robust? It's a lot more complicated than that, but predictions could have a useful role to play in science, and new projects are springing up to make use of them. ...  "

Numbers Being Transformations

Claimed success or sometimes not.  Rarely seen the actual numbers discussed.    Metrics are important.

The numbers behind successful transformations   By Kevin Laczkowski, Tao Tan, and Matthias Winter

“What gets measured,” Peter Drucker famously observed, “gets managed.” One might add a corollary that what goes unmeasured—or gets measured only superficially—risks being mismanaged or, at least, undermanaged.

So it is with transformations. As we’ve noted before, the term “transformation” can be vague, and it too often refers only to minor or isolated initiatives. What should define a transformation is in fact the opposite: an intense, well-managed, organization-wide program to enhance performance and to boost organizational health. And the results should always be measured.

Transformatics: Inside the metrics of transformation

As part of an analysis we term “transformatics,” we’ve built the capability to measure the data set we’ve assembled of more than 200 large transformations stretching back nearly a decade. More recently, we isolated the 82 public companies that had undertaken a full-scale transformation and had an observable 18-month transformation track record to see what we could learn from a statistical analysis of their experiences. The research highlighted four indicators that showed a statistically significant correlation with top-quartile financial performance during the 18-month test period (for more about the methodology, see sidebar “Transformatics: Inside the metrics of transformation”). Taken together, the four indicators suggest some potential lessons for senior managers seeking to maximize the odds of a successful transformation. Let’s look at each in turn.  ... " 

4 Keys to AI Success

Good, completely non technical piece in the Cisco Blog. Needs much filling out but starts the conversation.

AI and the 4 Keys to its Success: Diving Deeper
By Anthony McKinney, Cisco Blog
October 31, 2019 - 0 Comments

Like the internet, artificial intelligence (AI) becomes more powerful as the number of people using the tools grows. The utility of AI in this context depends on the user themselves. How? By providing their “voice samples” through applications like Siri, Alexa, Cortana and Google Assistant back to the developers. This helps build better algorithms. But these solutions need a solid foundation. One that supports the collection and the processing of the algorithms. But also an infrastructure that can adapt and support to scale, creating a robust solution for optimal utility.  ..." 

Wednesday, October 30, 2019

QR Codes Have Considerable Value

Someone recently mentioned to me, in a discussion of brand package identification, that QR Codes had disappeared.   I knew that was not the case, but did not have a good set of examples.  And then came this.  Here a very extensive look about how they are being used effectively in China.  Including illustrative pictures. Very useful if you are looking for examples.   I passed this along.

Remember QR Codes? They’re More Powerful Than You Think     by Avery Segal in Andreessen Horowitz

China’s mobile payment ecosystem, the largest in the world, is built upon QR codes. But that technology extends far beyond shopping to ease friction throughout daily life. On a recent trip to China, I personally interacted with QR codes 42 times in a single day—to ride the train, to book a workout, to charge my phone, even to buy a round of drinks for a stranger’s birthday.

Companies in the US have been slow to adopt QR codes, but those who dismiss them as having “been around forever but never taken off” underestimate their wide-ranging potential. Camera-based solutions like QR codes (or facial recognition, for that matter) can make traditionally clunky user experiences seamless and intuitive. QR codes connect our online identity to the offline world, allowing users to essentially log in to physical locations—and bring their data with them. This delivers a number of benefits: brands learn user preferences, while customers gain a more tailored and social experience, as well as perks like automatic loyalty programs built into every transaction.

It’s no coincidence that QR codes were popularized in China, where many consumers leapfrogged the PC and bought a smartphone as their first computer. As a result, many of China’s products are built first and foremost for mobile, a fact that was readily apparent on my recent trip to China. This list highlights just some of the ways QR codes enable a whole new set of mobile-first experiences: improving offline social interactions, promoting easy health monitoring, streamlining operational efficiencies, and establishing innovative shopping channels. Here are 16 (more!) ways China is using QR codes.  ... " 

Wind Energy Kites

Intriguing approach for wind energy production, in IEEE Spectrum.

Alphabet’s Makani Tests Wind Energy Kites in the North Sea
The flyers are just one of several airborne wind energy technologies that could shape the future of wind power    By Mark Anderson  ..... 

People-Centered Design Principles for AI

I attended the following, and I think the following links will work for you, but will test.

Watch your webinar on demand!

MIT Sloan Management Review

 ....  “People-Centered Design Principles for AI Implementation.” You can access the audio recording at any time.  Watch On Demand » https://t.e2ma.net/click/7f45wg/7znyzx/jt2ykkc

You can also download the slidedeck (PDF).  https://t.e2ma.net/click/7f45wg/7znyzx/zl3ykkc you are having trouble accessing the event materials, please contact us at smr-help@mit.edu.

Three People-Centered Design Principles for Deep Learning
Bad data and poorly designed AI systems can lead you to spurious conclusions and hurt customers, your products, and your brand. Read the article » https://t.e2ma.net/click/7f45wg/7znyzx/v64ykkc

Michael Goodkin on Getting Answers Faster

Recent reading inspires some thoughts.

Struck me that such methods what we are doing today.  Is faster always better?  And may be further advanced by technologies like 5G.   Depends on the risk involved and how it is managed.

The Wrong Answer Faster: The Inside Story of Making the Machine that Trades Trillions

(Michael Goodkin).   His company's original investment techniques became known as statistical and quantitative arbitrage. By 1996, these techniques accounted for most of the volume on the global exchanges and the financial derivatives market. Having resettled in Chicago, Goodkin then set out to make the market less risky by introducing computational physics to derivatives risk management.[1]

Goodkin’s most successful start-up was Numerix. Recruiting a group of academic physicists, including Mitchell Feigenbaum, winner of the MacArthur grant and the Wolf Prize in Physics for his pioneering work in Chaos Theory, Numerix was founded in 1996. The company’s initial product was a software algorithm that dramatically reduced the time required for Monte Carlo pricing of exotic financial derivatives and structured products. Numerix remains one of the leading software providers to financial market participants.[3]   .... "

Brain Scans Will Identify You

Not unexpected, with that much data.   But can it be done externally?  Methods are evolving in many kinds of biometrics.

Someday a Computer May Use Brain Scans to Identify You
The New York Times   By Gina Kolata

Mayo Clinic investigators said facial recognition software could be used to match photos of people to facial reconstructions derived from magnetic resonance imaging (MRI) scans of their heads. The University of Pennsylvania's Aaron Roth warned this technique eventually will be used to compromise stored medical data. The University of California, San Francisco's Michael Weiner said the Mayo Clinic's findings represent a threat to privacy, citing the Alzheimer's Disease Neuroimaging Initiative as a potential target. The Initiative has MRI brain scans that include participants' faces, with identifying data removed; Weiner suggested attackers could match those MRIs to images of study subjects elsewhere .... "

Notes and Introduction to Extreme Classification

Notes on. 
Extreme Classification   By Manik Varma  (Abstract)
Communications of the ACM, November 2019, Vol. 62 No. 11, Pages 44-45

What would you do if you had the super-power to accurately answer, in a few milliseconds, a multiple-choice question with a billion choices? Would you design the next generation of Web search engines, which could predict which of the billions of documents might be relevant to a given query? Would you build the next generation of retail recommender systems that have things delivered to your doorstep just as you need them? Or would you try and predict the next word about to be uttered by U.S. President Donald Trump?

The objective in extreme classification, a new research area in machine learning, is to develop algorithms with such capabilities. The difficulty of the task can be judged from the fact that, even if it were to take you just a second to read out a choice, it would take you more than 30 years to go through a billion choices. In 2012, state-of-the-art multi-label classification algorithms were struggling to pick the correct subset of options in questions involving thousands of choices. Then, in 2013, a team from Microsoft Research India and IIT Delhi developed a classifier1 that could scale to 10 million choices, thereby laying the foundations of the area. The approach was based on the realization that only a handful of choices would be relevant for any given question on average. The trick was therefore to quickly eliminate the millions of irrelevant choices. The classifier could then accurately and efficiently choose from the remaining hundred or so options. ..... "

Based on:
Extreme Classification - New Paradigm for Ranking and Recommendation

A good mostly non technical video presentation from 2018:   https://www.youtube.com/watch?v=n4jM5YnUQjE 

A Method for Ranking, Recommendation, Advertising applications,  Search, NLP, Vision...  Seen as replacing recommenders based on matrix factoring

Current methods started with Entropy based methods and lead to tree-splitting algorithms.

Tuesday, October 29, 2019

Talk: Sports Summary Highlight Video Construction Using AI

You can now find the recording from Stephen Hammer's excellent Oct 24th talk on improving the fan experience:    "Sports Summary Highlight Videos using AI"

Slides:    http://cognitive-science.info/wp-content/uploads/2019/10/csig_Sports__AI_Hammer_v1.120191024-2.pdf
Talk:      https://youtu.be/StDgf3mnKEU

#CSIGnews #opentechai #ISSIP #Wimbledon @usta #GRAMMYs @BaughmanAaron  @IBMWatsonMedia

Via Susan Malaika

Upcoming and past talks, given most weeks:   http://cognitive-science.info/community/weekly-update/ 

On the Current and Future Technology of Drones

Quite an interesting overview of the current technology of Drones.  Below the intro, article at the link.

When Drones Fly   By Samuel Greengard
Communications of the ACM, November 2019, Vol. 62 No. 11, Pages 16-18

As drones have matured into smarter and more practical machines, they have hummed, buzzed, and whirred their way into industries as diverse as movie production, agriculture, civil engineering, and insurance. It is entirely clear that autonomous drones will play a prominent role in business in the coming years. Firms such as Amazon, FedEx, and Uber have experimented with the technology to deliver packages, food, and more, while military agencies, emergency responders, gaming companies, entertainment firms, and others have explored other possibilities.

"Drones introduce far more efficient ways to accomplish some tasks," says Todd Curtis, president of Airsafe. com, a site that tracks drone and other aeronautic technologies.

Powering more advanced drones are more sophisticated on-board sensors and processors, better artificial intelligence (AI) algorithms, and more advanced controllers and communication systems. In addition, engineers are packing greater numbers of sensors into drones—and using them in different combinations—to create greater "awareness" of the surrounding environment. This sensing, when combined with GPS and other navigation capabilities, allows drones to tackle more advanced autonomous tasks, including devices that explore caverns or other hard-to-reach spaces, as well as underwater drones that conduct research by scanning oceans.

Yet, despite rapidly evolving capabilities, it also is clear that autonomous drones have not completely mastered the art and science of navigating and accomplishing their designated task. Buildings, birds, power lines, trees and people remain formidable obstacles for autonomous Unmanned Aerial Vehicles (UAVs), as they are known. Fog, snow, smoke, and dust present additional challenges.

It is one thing to showcase a drone in a controlled environment; it is quite another to have it operate flawlessly in the wild. UAVs must have near-perfect vision and sensing, as well as the ability to navigate areas where satellite and communications signals cannot reach and need backup and fail-safe systems that can take control of the drone if/when something goes astray.

"We are seeing remarkable advances in onboard sensing and processing, but also the use of far more sophisticated AI (artificial intelligence) algorithms in drones," says Nathan Michael, associate research professor at the Robotics Institute of Carnegie Mellon University. "These navigation and control systems are moving drones beyond the basic ability to fly from Point A to Point B. They're making it possible for drones to understand the world around them and make complex decisions in real time."  .... ' 

Vint Cerf on the 50th Anniversary of the Internet

Love the History of Science, here Vint Cerf, architect of the Internet, does a good job here of giving some important points its history.  Proud to have been there for the entire time.

Vint Cerf’s top moments from 50 years of the Internet  in The Google Blog .... Google VP and Chief Internet Evangelist

Editor’s note: On the 50th anniversary of the Internet, this post comes from one of the most knowledgeable sources out there. Though it’s not included in his official title, Vint Cerf is, in fact, one of the architects of the modern Internet. 

Before there was the Internet, there was a packet. The “sending of the packet” was actually the first step toward the invention of the Internet as we know it, and it happened 50 years ago today. On that day, we established the first connection between two computers—from UCLA to the Stanford Research Institute—on the ARPANET, the predecessor to the Internet. 

Connecting the planet in this way remains one of the most astounding technical and societal achievements of our lifetime. A lot has happened in the years since, and the rise of the Internet has come with its own set of challenges that will require new solutions. But over the years there have been many bright spots, including 17 moments that, for me, stand out the most. 

1. October 29, 1969: The first packet was sent. This pioneered our understanding of operational packet switching technology, which prepared us for the subsequent development of the Internet. 

2. 1971: Networked electronic mail was created using file transfers as a mechanism to distribute messages to users on the Arpanet.

3. 1974: The design of the Internet was released. Robert Kahn and I published “A protocol for packet network intercommunication.” In this paper we presented not only a protocol, but an architecture and philosophy that supported an open design for the sharing of resources that existed on different packet-switching networks. 

4. November 22, 1977: A major demonstration of the Internet took place, linking three networks: Packet Radio, Packet Satellite and ARPANET. 

5. January 1, 1983: The Internet was operationally born, and I’ve used an “electronic postcard” analogy to explain how it works.

6. 1983: The operational mobile phone arrived, which is crucial because, although the Internet and mobile phones were developed in parallel, they eventually proved to be complementary technologies. ... "

  (Read the rest at the link!)

Interview with Alibaba Group Chairman and CEO

On innovation in analytics in particular, like the use of the term 'analytics' rather than always lapsing into AI.   AI is only one way to do and deliver analytics.

Speak softly, make tough decisions: An interview with Alibaba Group chairman and CEO Daniel Zhang.    The chairman and CEO of China’s e-commerce giant describes Alibaba’s approach to innovation and how he balances analytics and instinct to push himself to spot hidden opportunities.

Sent from McKinsey Insights 

Ikea Plans a Smart Button for the Home

Am a fan of the idea of touch as well as voice.   Like in the now defunct Amazon Dash.   Now is nascent smarthome constructor IKEA taking over this channel?

Ikea’s smart button leaks with a tease of scenes to come
Ikea’s inexpensive Home smart ecosystem continues to grow

By Thomas Ricker in TheVerge

Ikea’s working on a new Shortcut Button that can activate a “scene” in homes that are fitted with the company’s smart products, according to a new Federal Communications Commission filing. A scene is typically defined as a set of event-driven (leaving home, having dinner, etc.) commands that are issued simultaneously to multiple smart home devices. The Tradfri Shortcut Button was first spotted by Dave Zatz, with Swedish site Teknikveckan speculating that it might be possible to add different pictograms behind the button’s plastic door to better identify its function.... "

Splunk and the Data Problem

Had several interactions with Splunk, in the enterprise,liked it for streaming data problems.  But I would rather say that every problem has an embedded data problem, and that size of that can vary according to the business context. Don't remember Splunk being particularly aimed or useful at that, but seems that are building towards that.  Will take a look.

Every problem’s a data problem, says Splunk. Can its new platform fix them?  By R. Danes in Siliconangle

Splunk Inc. is getting serious about this data platform thing. The company wants to get friendly with data from any source — not just the Splunk index. The idea is that a large, inclusive platform can ultimately get more juice from data — business insights, social impact, etc. — than a hodgepodge of software products.

“We believe, at the heart of every problem, is a data problem,” said Susan St. Ledger (pictured), president of worldwide field operations at Splunk. Think that’s an overstatement? St. Ledger named wildfires, the opioid crises, and human trafficking as examples of the issues people are attacking with data today.  ..." ... '

Researching the Open Office

Good to see actual research being done in the space.   Depends much on personal habits and previous experience,  have seen it both ways.

Why Open Offices Aren’t Working — and How to Fix Them in the HBR

Ethan Bernstein, associate professor at Harvard Business School, studied how coworkers interacted before and after their company moved to an open office plan. The research shows why open workspaces often fail to foster the collaboration they’re designed for. Workers get good at shutting others out and their interactions can even decline. Bernstein explains how companies can conduct experiments to learn how to achieve the productive interactions they want. With Ben Waber of Humanyze, Bernstein wrote the HBR article “The Truth About Open Offices.” .... "

Crafting Persuasion

Written by some of my former colleagues,  looks to be quite good.    There is some useful content at the link, as well as pointers to buying the book.  This is useful even if you do have formal training in communications.

Handbook to Change Minds and Influence Behavior - Crafting Persuasion
Former Senior “P&G Plus” Executive Co-Authors A Must-Read Book – Crafting Persuasion

Every leader realizes the importance of an effective communication strategy. But, how do you consistently succeed in telling the story of your brand or organization? It takes a model and a guidebook written by authors who have spent decades learning the art and science of creating powerful stories for some of the world's best marketing organizations. Crafting Persuasion describes the ABCDE model (audience, behavior, content, delivery and evaluation) in a step by step guide that is as important to storytelling as the 4 P's (price, product, promotion and place) are to marketing. It is a model that works in any setting. 

Whether you are an engineer running a start-up or an NGO trying to sway public opinion or a CEO evaluating a marketing plan, or a government officer charged with communicating a critical message or policy, this book is for you. The strategic principles of Crafting Persuasion have been taught over the past decade at the U.S. State Department and other leading organizations. It is designed for those who have the responsibility to persuade an audience, but never had any formal training on how to do this. It is filled with real-world examples from the business and non-profit worlds, along with an enlightening companion website (www.craftingpersuasion.com). So, regardless of your communication challenge, Crafting Persuasion will show you how to create a communication strategy to win over audiences and reach your personal and professional goals.  .... " 

McDonald's Use of Sales AI

Note the use of sensor acquisition to drive sales category with AI methods.  Lots of data there, and many different kinds of components.   Have always respected McDonald's tech work.

Would You Like Fries With That? McDonald's Already Knows the Answer
The New York Times
By David Yaffe-Bellany

McDonald's is acquiring companies that develop artificial intelligence and machine learning to make the company more like Amazon. Its incorporation of technology is aimed at reversing its recent loss of customers, resulting in restaurants closing and sales declining. The company is incorporating technologies such as digital boards that promote its products, taking into account environmental factors like the weather and the length of the wait for service. The company has tested algorithms at its drive-throughs that capture license-plate numbers, so the restaurant can list recommended purchases personalized to a customer's previous orders, as long as the person agrees allow the fast-food chain to store that data. McDonald's also recently tested voice recognition at certain outlets, with the goal of deploying a faster order-taking system. Regarding the use of new technologies, the company’s CIO, Daniel Henry, said, “You just grow to expect that in other parts of your life ... We don’t think food should be any different than what you buy on Amazon."  ... '

Monday, October 28, 2019

Contextual Intelligence: A Next Big Thing

Good overview piece in DSC.   I add some of our own experience using AI techniques.

We struggled with this a number of times.   All assistance exists in some context.  The context itself can be complex or not.   So for example we built a pump replacement system for choosing industrial pumps based on a dozen criteria.   Though this was technical, even including some predictive analytics,  it was straightforward.    Also easy to explain its details to the decision makers involved.   It also produced some additional meta information:  " how much did we, and would we spend on XYZ pumps next year, and what should their maintenance cost be?

But once we looked at complex systems.  Ones that depended on human behavior, required common sense decisions,  competitive reaction, this needed many sensory inputs, depended on the decisions of many other people and systems, the whole thing became difficult.  All this was asked in questions like "What profit is the company likely to make on new product X in the next year?"   At one point we estimated that such a question required several hundred external inputs.  Assistance in complex context is hard.  And you always end up with some risk/uncertainty attached to the assistance.     Even how the decision is implemented can add considerable risk and uncertainty.   Much work to be done.

Contextually Intelligent NLP Assistants – AI’s Next Big Technical Challenge   Posted by William Vorhies 

Summary:  Contextually intelligent, NLP-based interactive assistants are one of the next big things for AI/ML.  The tech is already here from recommendation engines.  The need to be more efficient and to become AI-augmented in our decision making is now.  Getting the contextual awareness is the hard part.

Last week we took the position that from a technical standpoint, ‘deeply inclusive and contextually sensitive’ AI is one of the two ‘next big things’ in AI.

In retrospect I wish there were a more concise agreed naming convention for this bit of technical legerdemain.  “Inclusive” and “contextually sensitive” are in the category of those ‘suitcase words’ Marvin Minsky called out as being so dependent on the user’s experience that agreement on meaning is difficult.

What we’re not talking about is the ability of NLP to hold a contextually appropriate conversation, such as making a reasonable response or request for clarification based on the topic at hand.  For the most part, short of performing psychoanalysis, chatbots can do pretty well with human ad hoc conversation.

Also, we’re not talking about being culturally inclusive as in detecting and eliminating bias.  Important, but not what we’re getting at.

What we’re describing is the next big step in NLP utility in which the NLP puts together facts it knows about us and proactively takes action or makes suggestions that make our life easier.

The example we gave in our previous article is about having the NLP assistant remind me of my mother’s upcoming birthday in a week or so without my having explicitly created a reminder.  More importantly my NLP assistant could make a recommendation for a present.  Presumably my past communications with her both in fact and tone contain some strong signals about my mom’s demographics and perhaps even her interests so why not predict a short list of appropriate gifts.  Now that would be valuable.

So perhaps a better description of this behavior then would be ‘contextually intelligent’.  We’ll stick with that.

Smart Carts aiding Self-Checkout

We spent much time examining and testing the idea of a 'smart cart', but none of them ever became commercially viable.    See my coverage at the tag.  Hand held dedicated devices became the replacement, and later the no-checkout idea promoted by 'Amazon Go'.

Tired of Long Lines? Canadian Grocery Chain Debuts Smart Carts with Self-Checkout   The Washington Post   By Peter Holley

Nova Scotia, Canada-based grocery chain Sobeys has launched a pilot program using intelligent shopping carts that scan and weigh items, and help customers skip long checkout lines by allowing them to pay on the spot. Sobeys' Smart Cart fleet features touchscreens that display a running count of purchases as shoppers scan and place their items in bags within the cart; customers can pay as soon as their shopping is completed. Sobeys says that as the carts are upgraded, their screens will help customers navigate stores, fill out shopping lists, and suggest products for recipes. The carts are equipped with high-resolution cameras which, when combined with scales, enable shoppers to add items to their purchase without entering information or scanning bar codes. ... "

Drones are Filling the Skies

A broad piece in the WSJ on how major players are examining, testing and using drones.  And the methods they are using to make this all work.

The Drones Are Coming! How Amazon, Alphabet, Uber Are Taking to the Skies
The Wall Street Journal
By Sebastian Herrera; Alberto Cervantes
October 25, 2019

Companies including Amazon, Alphabet's Wing, and Uber, are launching more advanced trials of drone delivery. Wing started tests in Christiansburg, VA, this month, while Uber will set up experiments in San Diego before the end of the year. Amazon said last June it would begin delivering packages to consumers via drone "within months." The companies have to overcome a number of obstructions and concerns before drone delivery can become widespread. Amazon uses machine learning algorithms and infrared sensors to detect obstacles like birds and wires, and programs its drones with scenarios (such as when a delivery location cannot be detected), and commands to follow in such scenarios. Wing, meanwhile, has tested its drone north of Helsinki under snowy and windy conditions; its drone has built-in wind sensors and is waterproof. A challenge that remains is that no standard exists on how drones can identify and communicate with each other while in flight, so drone delivery by multiple companies in the same area is not currently possible. ... ' 

TODAY: Computer Science and Law

TODAY, of interest:

Watch Livestream of ACM Symposium on Computer Science and Law

To view image click on Watch a livestream of the inaugural ACM Symposium on Computer Science and Law being held at the New York Law School in New York City today from 8:00 am to 7:00 pm EST. Click on the “Watch the Event Live Stream” button on the event website to view invited talks by speakers including keynoters Shafi Goldwasser (2012 ACM Turing Award co-recipient), Edward Felten (ACM Fellow and former ACM US Public Policy Council Chair) and Jack Balkin (Knight Professor of Constitutional Law and the First Amendment at Yale Law School). Panels will explore research, education, and practice in the interplay of computer science and law. The day will conclude with a reception featuring student posters about work in computing and law, on topics ranging from cybersecurity to legal informatics.

Click here for the full-day program.  https://computersciencelaw.org/program 

A Talk on Communal Intelligence

We have not even gotten modeling individual intelligence right yet.  And it is that intelligence that has provided so much striking value.  So do we want to be 'communal' quite yet?   Cooperative and effective yes, but communal?   Squelching the individual spirit?

Communal Intelligence    A Talk By Seth Lloyd 

We haven't talked about the socialization of intelligence very much. We talked a lot about intelligence as being individual human things, yet the thing that distinguishes humans from other animals is our possession of human language, which allows us both to think and communicate in ways that other animals don’t appear to be able to. This gives us a cooperative power as a global organism, which is causing lots of trouble. If I were another species, I’d be pretty damn pissed off right now. What makes human beings effective is not their individual intelligences, though there are many very intelligent people in this room, but their communal intelligence.

SETH LLOYD is a theoretical physicist at MIT; Nam P. Suh Professor in the Department of Mechanical Engineering; external professor at the Santa Fe Institute; and author of Programming the Universe: A Quantum Computer Scientist Takes on the Cosmos. Seth Lloyd's Edge Bio Page .... 

Standards for Robotics in Retail

An interesting document on standards for robotics in retail:

via .... Platt Retail Institute (PRI) is an international consulting and research firm that focuses on leveraging technology to impact the consumer experience and store operations. Central to this is building actionable data models that aid retailers and technology companies in gaininginsights into their customersand operations. In addition to its global consulting expertise, PRI also publishes pioneering industry research  ... 

Standards Support Robots in Retail
By Richard Halter, President, Global Retail Technology Advisors, LLC .... '

Vehicles Seeing Around Corners

Interesting advances continue in the sensor space.  Another example of sensors inferring related information.

Helping autonomous vehicles see around corners
By sensing tiny changes in shadows, a new system identifies approaching objects that may cause a collision.   Rob Matheson | MIT News Office

To improve the safety of autonomous systems, MIT engineers have developed a system that can sense tiny changes in shadows on the ground to determine if there’s a moving object coming around the corner.  

Autonomous cars could one day use the system to quickly avoid a potential collision with another car or pedestrian emerging from around a building’s corner or from in between parked cars. In the future, robots that may navigate hospital hallways to make medication or supply deliveries could use the system to avoid hitting people.

In a paper being presented at next week’s International Conference on Intelligent Robots and Systems (IROS), the researchers describe successful experiments with an autonomous car driving around a parking garage and an autonomous wheelchair navigating hallways. When sensing and stopping for an approaching vehicle, the car-based system beats traditional LiDAR — which can only detect visible objects — by more than half a second.  ... "

Contract Management with AI and RPA

Have seen a number of contract management approaches now, from the very simple to the complex.  Make sense to start simple and progress based on goals, needs, and risks.

Contract Management 2.0: why AI and RPA will boost outsourcing results

The world is changing in a fast pace due to the overwhelming developments in Technology and operational models. Just like any market, also Contract management is subject to the influences of digital transformation. Technologies like artificial intelligence and robotic process automation are slowly entering the Contract management market. How will contract automation impact contract management? Technology evangelist Arjen van Berkum wrote an interesting article on these topics. We are happy to share his vision with our readers.

Overall, Technology is improving the way we work. When it comes to managing contracts these improvements – mostly contract automation related – have a positive impact on  business results, supplier management as well as on customer and employee satisfaction. Technologies like artificial intelligence (AI), robotics and blockchain can improve contract management issues like compliance, data management, invoice settlement and contract analysis.

Technology does not need any sleep, it is not bound to a nine to five mentality and it doesn’t get sick. It offers a 24/7 availability and, when implemented well, it is more accurate than humans can ever be as long as it is bound by rules and repetition. As like almost everything in our professional lives, contract management is being influenced by digitization. Contract managers are changing their way of work and are eager to embrace the technological developments surrounding them. This is only logical, because technology carries a lot of potential in the contract management field, and I will explain how.   ... " 

Sunday, October 27, 2019

Data Management for Data Science

In Kdnuggets a good description and visualization of data management needed for data science.

Everything a Data Scientist Should Know About Data Management

For full-stack data science mastery, you must understand data management along with all the bells and whistles of machine learning. This high-level overview is a road map for the history and current state of the expansive options for data storage and infrastructure solutions. By Phoebe Wong and Robert Bennett.

To be a real “full-stack” data scientist, or what many bloggers and employers call a “unicorn,” you have to master every step of the data science process — all the way from storing your data, to putting your finished product (typically a predictive model) in production. But the bulk of data science training focuses on machine/deep learning techniques; data management knowledge is often treated as an afterthought. Data science students usually learn modeling skills with processed and cleaned data in text files stored on their laptop, ignoring how the data sausage is made.   ... " 

Tiny Drones for Unknown Environments

We explored the approach of using small swarms of drones.   In industrial and outdoor forestry and logging spaces.....

Swarm of Tiny Drones Explores Unknown Environments
Delft University of Technology

A swarm of tiny drones that can explore unknown environments on their own was created by researchers from the Netherlands’ Delft University of Technology (TU Delft) and Radboud University of Nijmegen, and the U.K.'s University of Liverpool. The camera-equipped, 33-gram drones navigate autonomously, with limited sensing and computational capabilities. A proof of concept for search-and-rescue operations demonstrated that a six-drone swarm could investigate about 80% of open rooms, and swarming added redundancy so data otherwise lost by one malfunctioning drone could be supplied by others. A wireless communications chip installed in each drone allows them to detect and avoid each other by reading signal strength between the chips. Delft’s Kimberly McGuire said, “The main advantages of this method are that it does not require extra hardware on the drone and that it requires very few computations.”

Giant Food for Frictionless Checkouts

There has been some suggestion that Amazon is backing off the idea, except in narrow contexts.  But recently there has been a newly opened store.  Here new word in the spec from Ahold, who we worked with  on several advanced tech applications, though not this one.   This includes specific description of the process, which is different than I have seen.

Have Giant Food and Stop & Shop nailed ‘frictionless’ checkouts?  includes further expert comments.   In Retailwire  By George Anderson

Retail Business Services, the technology services arm of Ahold Delhaize USA, has announced that its proprietary ScanIt Mobile frictionless checkout technology is being rolled out to nearly 30 of the grocer’s stores by the end of the year. The tech is being deployed at all of the new Giant Heirloom Markets and select Stop & Shop stores.

Customers using the ScanIt mobile app walk shop the store scanning products they wish to purchase as they go. When finished, customers go through a designated checkout lane where they see a “payment approved” message before exiting the store. Payments are processed through customers’ mobile wallets. The service accepts Apple Pay, Google Pay, PayPal and Venmo.

Individuals who prefer may also use handheld ScanIt devices supplied by the store. Here too, they scan products as they move around the store before transferring the information to the mobile app on their phones to check out. ,,, "

Decade's end: On Future Developments

 Gizmodo does a reasonable look at what to expect:

The Most Futuristic Developments We Can Expect in the Next 10 Years
By George Dvorsky in Gizmodo  via Walter Riker

Decade's End

Gizmodo, io9, and Earther look back at our passing decade and look ahead at what kind of future awaits us in the next ten years.

With the decade winding down it’s time for us to set our sights on the next one. The 2020s promises to be anything but dull. From the automation revolution and increasingly dangerous AI to geohacking the planet and radical advances in biotechnology, here are the most futuristic developments to expect in the next 10 years.

Making predictions is easy; it’s getting them right that’s tough. That said, some tangible trends are emerging that should allow us to make some informed guesses about what the future will hold over the next 10 years. .... " 

China Passes Cryptography Law

Broad implications unclear, especially how it might relate to Blockchain or other uses of cryptography that relate to regulatory access.  Apparently quotes the Chinese president promoting blockchain.

China’s Congress Passes Cryptography Law, Effective Jan. 1, 2020 In Coindesk

The Standing Committee of the 13th National People’s Congress in China passed a cryptography law on Saturday that will be effective on January 1, 2020, according to a Chinese media report.

The announcement came one day after Chinese President Xi Jinping called on the country to seize opportunities in blockchain technology.

While China still bans cryptocurrency trading and its national digital currency is not yet hatched, cryptography, as an integral underpinning of blockchain technology, could be key to the country’s push to be more competitive in the blockchain space.

The new law aims to tackle emerging regulatory and legal challenges in commercial cryptography use-cases as they play an increasingly important role in developing the Chinese economy, according to the law’s latest draft proposal prior to approval.

According to the proposal:

“Clear guidelines and regulations are needed to evaluate commercial cryptography technologies used in the major fields related to the national interest as the current ‘loose’ system is not suitable for the industry anymore.”  .... " 

Saturday, October 26, 2019

Decision Trees for Shopping

Perhaps simplistic, but interesting.   Even if you have to ultimately complicate such a model, its useful to start with something this simple and see how well it can predict, then add more complexity later.  In the enterprise, we did lots of that.  It can also act as a benchmark for more complex models.  Decision trees are also nicely transparent.

Decision Trees for Online Shopping Analysis
Towards Data Science by Chathuranga Siriwardhana  

Nowadays there is a trend to use online shopping solutions like Amazon, eBay, AliExpress. These websites provide a platform for the sellers to sell their products to a large number of customers. Since many delivery services are connected with these online shopping platforms, customers from different countries buy products. Unlike the traditional shops, the ratings and the good-name is directly represented on the shopping platform for each seller. Therefore the sellers have let the customers return their bought items if they don’t like the product or there is any defect of the item. Some sellers refund the whole amount if the customers complain that the items are not delivered within the promised period. Some customers are misusing these facilities and fraud to the sellers. Therefore, the sellers on the online shopping platforms experience a huge loss of profits. Let’s discuss how we can spot these types of customers by developing a simple Machine Learning model; a Decision Tree.

Have a look at this medium post on Decision Trees if you are not familiar with them. For a quick recap, a decision tree is a model in machine learning which includes the conditions on which we are categorizing the data (for labelling problem). As an example, think about a simple situation where a man is happy is the weather is sunny or he is on vacation. This scenario is modelled below. Note that you can use weather and vacation status to predict the man’s happiness with this model.  ... "

Cross Entropy

What is this?  New to me as a term.  I record here for my own refernce and pass it out to others.   In particular the suggestion here is that classification may be improved or even optimized this way.   See Jason's other publications, some mentioned here.

A Gentle Introduction to Cross-Entropy for Machine Learning   by Jason Brownlee

Cross-entropy is commonly used in machine learning as a loss function.

Cross-entropy is a measure from the field of information theory, building upon entropy and generally calculating the difference between two probability distributions. It is closely related to but is different from KL divergence that calculates the relative entropy between two probability distributions, whereas cross-entropy can be thought to calculate the total entropy between the distributions.

Cross-entropy is also related to and often confused with logistic loss, called log loss. Although the two measures are derived from a different source, when used as loss functions for classification models, both measures calculate the same quantity and can be used interchangeably.

In this tutorial, you will discover cross-entropy for machine learning.

After completing this tutorial, you will know:

How to calculate cross-entropy from scratch and using standard machine learning libraries.
Cross-entropy can be used as a loss function when optimizing classification models like logistic regression and artificial neural networks.
Cross-entropy is different from KL divergence but can be calculated using KL divergence, and is different from log loss but calculates the same quantity when used as a loss function.
Discover bayes opimization, naive bayes, maximum likelihood, distributions, cross entropy, and much more in my new book, with 28 step-by-step tutorials and full Python source code.

Let’s get started: ... " 

Friday, October 25, 2019

Qualcomm Creates Fund for 5G Use Beyond Phones

Being speed, what are the applications in play?  Where can we predict particular useful ventures that we should support?

Qualcomm creates $200 million fund for 5G uses beyond phones in Reuters

Qualcomm Inc on Thursday said it has created a $200 million venture capital fund to invest in startup companies looking to use 5G technology in devices other than smartphones.

(Reuters) - Qualcomm Inc on Thursday said it has created a $200 million venture capital fund to invest in startup companies looking to use 5G technology in devices other than smartphones.

Qualcomm is supplying chips for 5G, the next generation of mobile networks rolling out this year, to phone makers such as Samsung Electronics Co Ltd The San Diego company is the biggest supplier of chips for mobile phones, but that market has stopped growing.

Qualcomm is looking to invest in companies that will use 5G in other areas, such as connecting industrial and agricultural equipment to the internet or enabling self-driving cars to communicate with infrastructure such as traffic lights and signs.

“5G will transform industries and should be viewed as a business strategy for all,” Steve Mollenkopf, Qualcomm’s chief executive, said in a release.  .... " 

Wozniak: No Self-Driving Cars in My Lifetime

A contrary look to be considered.

Steve Wozniak: No Self-Driving Cars in My Lifetime  By Bill Howard  in Extremetech

LAS VEGAS — Apple co-founder Steve Wozniak believes in technology. But that doesn’t extend to believing autonomous driving is happening soon. Wozniak, now 69, says autonomous cars that don’t need a backup driver on board probably won’t happen “in my lifetime.” One culprit: Artificial intelligence probably isn’t intelligent or flexible enough to be better than even the worst drivers.

Wozniak was a keynote speaker at the first J.D. Power Auto Revolution conference, which the company set up to  “fuel innovation and drive an auto revolution,” with a bit less emphasis on the how-to of selling and marketing cars than some other Power programs.

Research cited at the conference said consumers see many forms of self-driving happening in eight to nine years, while experts see a spread of roughly five to 15 years, with the most distant goal being full autonomy where the driver doesn’t need to be always ready to take over for the car. One takeaway might be that consumers just have no idea how close we are to full self-driving and make the same timeframe guess no matter what level of automation is being considered, where technologists and engineers see some autonomy being harder than other types.   ... " 

Using Databricks with Tableau

Intriguing update for Tableau.

Visualize your data lake with the new Tableau Databricks Connector

Tableau 2019.3 was a momentous release for a number of reasons. Along with the unveiling of Tableau Catalog and Explain Data came a new native connection to Databricks for Tableau Desktop and Tableau Server. The new connector offers better performance, a straightforward connection experience, and high-quality error handling.

At Tableau, we’re thrilled to partner with Databricks to empower the full spectrum of data people throughout an organization. Databricks is helping data teams solve the world’s toughest problems, while Tableau makes it fast and easy to connect, explore, and make decisions data-driven. The two platforms are on a mission to make data more accessible and to enable organizations with self-service analytics, so creating a finely-tuned connector was an obvious next step for the partnership. This native connection is intended to better serve our customers as their organizations scale and their data strategies evolve.  .... "

A Phone Made of Paper

A cute and interesting idea.   Its best understood by looking at the very short video at the Fastcompany article below.  Makes you think about how we are processing our interaction with the world. 

Google just released a ‘phone’ made of paper
All you need to go offline for a day is a printer.
By Mark Wilson in Fastcompany

Our phones are making us unhappy, and companies like Google are wrestling with how to keep expanding a product like Android without destroying the soul of humanity in the process. The company’s designers have been outspoken on the topic, and they released a series of digital wellness tools to help track and manage phone use last year.

Now, the company is going a step further. It’s proposing that you replace your Android phone with a paper one—at least for a day.   ... " 

Advances in Conversational Search

Here a long time interest in intelligent conversation, and one of the most used uses of conversation on line these days is search.   Intelligent conversation gets us much closer to general AI.    Has this now been considerably improved?  Are we getting much closer, or is this one more demo that cannot be broadly delivered?   Note too how Google mentions people adapting to machines with 'keyword-ese'  and avoiding natural conversation. 

Google now understands more conversational search queries
The tech giant says it's one of the biggest Search updates in the product's history.
By Mariella Moon, @mariella_moon  in Engadget ..... 

Understanding searches better than ever before
Pandu Nayak in the Google Blog
Google Fellow and Vice President, Search

If there’s one thing I’ve learned over the 15 years working on Google Search, it’s that people’s curiosity is endless. We see billions of searches every day, and 15 percent of those queries are ones we haven’t seen before--so we’ve built ways to return results for queries we can’t anticipate.

When people like you or I come to Search, we aren’t always quite sure about the best way to formulate a query. We might not know the right words to use, or how to spell something, because often times, we come to Search looking to learn--we don’t necessarily have the knowledge to begin with. 

At its core, Search is about understanding language. It’s our job to figure out what you’re searching for and surface helpful information from the web, no matter how you spell or combine the words in your query. While we’ve continued to improve our language understanding capabilities over the years, we sometimes still don’t quite get it right, particularly with complex or conversational queries. In fact, that’s one of the reasons why people often use “keyword-ese,” typing strings of words that they think we’ll understand, but aren’t actually how they’d naturally ask a question. 

With the latest advancements from our research team in the science of language understanding--made possible by machine learning--we’re making a significant improvement to how we understand queries, representing the biggest leap forward in the past five years, and one of the biggest leaps forward in the history of Search.   .... " 

Recovering “Lost Dimensions” of Images, Video

Fascinating idea I could have used long ago.  Projections, for example, from advertising copy, archived and needed for reapplication. 

Recovering “Lost Dimensions” of Images, Video
MIT News   Rob Matheson

Researchers at the Massachusetts Institute of Technology (MIT) have developed a model that recovers valuable data lost from images and video that have been "collapsed" into lower dimensions. Captured visual data often collapses data of multiple dimensions of time and space into one or two dimensions called "projections." The researchers created a "visual deprojection" model that uses a neural network to learn patterns that match low-dimensional projections to their original higher-dimensional images and videos. The model takes in new projections and uses what it has learned to recreate the original data. During testing, the model synthesized accurate video frames showing people walking by extracting information from single-one-dimensional lines. The model also recovered video frames from single, motion-blurred projections of digits moving around a screen. ..."

MIT Connecton Science and Human Dynamics Lab

These efforts well worth following.   AI, Data, Human Dynamics, Privacy

MIT’s Pentland Outlines Rules For Data And AI
October 25, 201979
MIT Connection Science and Human Dynamics labs     https://connection.mit.edu/ 

There are several steps you can take using AI to ensure your data is secure.     By Allison Proffitt in AITrends

BOSTON—AI is data; data is AI. They’re really the same, Alex Pentland told a packed opening plenary session on the second day of the AI World Conference and Expo. “The winner is the person who has the most data. That’s probably not you, but you have friends who have data. They probably aren’t going to just give it to you; you have to figure out how to collaborate.”

Pentland directs the MIT Connection Science and Human Dynamics labs. He’s working on getting AI into the mainstream “peacefully”, he said. That is, no riots in the street, no mass unemployment, no cyber security onslaughts.

His group at MIT, sponsored by Ernst and Young, IBM, MasterCard, Orange, and others, builds pre-standards open source code—Kerberos, the network authentication protocol, for example—and tackles the big questions: How to ensure compliance, control risk, ensure privacy, and security? “Privacy is coming,” he warned, “not just in Europe and California, but everywhere.” ... "

Thursday, October 24, 2019

Predicting Molecule Smell

This sounds a bit odd, but we were trying to exactly this some time ago, when were still in the coffee business.  Smell, taste, and other sensory aspects of products, in both their blended and growing characteristics.  And of course the fragrance industry might also be interested. 

Google is training an AI to predict a molecule’s smell
by Ivan Mehta in ThenextWeb

With plenty of mics and cameras at disposal, AI has gotten good at ‘seeing’ and ‘listening.’ But one human sense it hasn’t got around much is smell. Now, researchers at Google are trying to develop a neural network that helps an AI identify the smell characteristics of a molecule.

The company said identifying smell is a multi-label classification problem, meaning a substance can have multiple smell characteristics. For instance, Vanillin, a substance often used to create an artificial vanilla flavor, has multiple smell descriptors such as sweet, vanilla, and chocolate, with some characteristics stronger than others.   ....'

Upcoming Developer Conference

I plan to attend the following free, online conference:

Upcoming AI conference

The Digital Developer Conference: AI & Cloud is designed for developers interested in Cloud and AI technologies by addressing the unique needs of coders. At this free online conference, get hands-on experience and engage with expert developers who will share insights on topics ranging from AI/ML innovation from IBM AI Research, open-source deep learning, model bias identification, multicloud best practices, cloud security with DevSecOps, and getting the most out of cloud native development. See client stories and how technology is being deployed to address some of the biggest issues facing developers today.

Important: The registration flow requires that you continue on to our resources page and accept our terms and conditions before you’re officially registered for the conference. In addition, if you already have an IBM Cloud account, you can skip this step and go right to the "Get Started" challenge on the resources page to complete your registration.

November 2, 2019 in North America   (For other regions/times see at the link)   .... " 

Is 5G Ready for IOT?

Makes sense to look at applications in the IOT space.

Is 5G Ready for IoT?

Prashant Gurav - views in Customerthink.

 5G has finally arrived but the real question for those looking to hire IOT development services is, Is 5G viable for IoT applications? We’ll determine through advantages and disadvantages.

5G has finally arrived and consumers have started to get a glimpse of it in the first world countries like the US, UK and South Korea. Being the first network that is built with IoT in mind, 5G is expected to make a big difference in the long term. As for those looking to hire IOT development services for their projects, the question remains, Is 5G viable for IoT applications?

Advantages of 5G for IoT
It is indeed surprising to know that some of the big advantages of 5G regarding consumer applications like high connection speed and greater data capacity aren’t relevant when it comes to IoT because IoT uses a large number of devices sending small amounts of data. Although theoretically speaking, this extra capacity could be a boon for IoT app development but practically there is no case where the density of devices has been too much for any existing network.

When we say 5G was built keeping IoT, we are not talking about the high connection speed or greater data capacity, on the contrary, it is the low power usage of this technology that hits home. The previous existing technology was built with an assumption that the main users will be using mobile phones which have high battery reserves and are needed to charge only once a day. While one can notice that seeing the increasing battery capacity of each mobile phone generation, but 5G is different here. 5G is much better when it comes to optimization which means it is much better equipped to deal with devices sending small amounts of data which in turn reduces the overhead of signalling and the payload for any particular bit of data.

Either way, the benefits of a lower battery consumption technology model of 5G are there to see in case of usage in IoT because devices with lower battery capacity can provide higher flexibility in how an IoT device is deployed while devices with high battery capacity can be left in the field for longer durations of time without needing maintenance.  ..... '

Augmenting Business Productivity with Digital Workers

Like the idea of taking RPA methods are defined closer to the idea of augmenting AI.   In our experience this is where applications of AI tech worked best.   Links to an interview:

IBM aims to increase business productivity with ‘digital workers’  By Silvia Fregoni in SiliconAngle

There is no doubt that automation technology helps enterprises increase scale and productivity. But there is always room for improvement. To leverage business results with automation, IBM Corp. is increasingly focused on using AI in this process and, as part as of its Cloud Pak for Automation, has announced a new capability called “digital workers.”

“The idea is that you can leverage a digital workforce,” said Mike Gilfix (pictured), vice president of digital business automation at IBM. “You can manage them like people, they can work alongside your people, and they can help to free up your people to be much more productive.”

Gilfix spoke with Dave Vellante (@dvellante), host of theCUBE, SiliconANGLE Media’s mobile livestreaming studio, during the IBM Data and AI Forum in Miami, Florida. They discussed new trends and challenges for business automation (see the full interview with transcript here). (* Disclosure below.)

To develop this new application, IBM has partnered with robotic process automation vendors. While RPA is more dedicated to repetitive and mundane tasks, digital workers can perform more critical and complex assignments, according to Gilfix.

“If you really look at where RPA is making its strides today, it is in data entry and sort of automation of input and data,” he explained. “The digital worker works just like a person does. It can sift through documents to find out what to take action on, help with decision-making processes, figure out when to act, how to prioritize work, and it can integrate into those people’s workflow.”  ... " 

Exploiting Multi-Categorical Features

Had not heard of this in particular, have passed it on.

Exploiting Multi-Categorical Features Using Deep Interest
By Marina Gandlin | Data Science
Tags: algorithms, big data, data, data model, data science, deep learning, machine learning, neural networks

At Taboola, our goal is to predict whether users will click on the ads we present to them. Our models use all kinds of features, yet the most interesting ones tend to be related to the users’ history. Understanding how to use these features well can have a huge impact on the model’s personalization capabilities, due to the user-specific knowledge they hold.  ..... ' 

More Amazon Go

Another example of extending the idea.  Mostly in small convenience formats due to costs.

Silicon Valley Takes on a New Amazon Cashierless Store

The Wall Street Journal
By Sebastian Herrera

Amazon recently opened its fourth cashierless Go convenience store in San Francisco, located within a few blocks of the other three. Nearby, two startups are each demonstrating their own technology that could power cashier-free stores across the country. This area of San Francisco is emerging as a battleground to eliminate the traditional checkout process and reinvent the way consumers shop. Amazon, along with startups Zippin and Standard Cognition, use technology equipped with camera systems powered by computer vision and machine learning software that track people as they take items off the shelves. The companies are pitching their systems to grocery chains, sports stadiums, and convenience stores, promising to automate the checkout process, reduce theft, and improve profit margins. The technology is currently only being tested in small convenience-store concepts selling packaged goods, because it is relatively expensive for a big-box retailer to adopt such systems on a wide scale. ... " 

Wednesday, October 23, 2019

Quantum Supremacy

Further commentary on Quantum Computing, by the CEO of Microsoft. Not only about the supremacy claim, but also in general about quantum computing and what it will mean in the future.

Google CEO Sundar Pichai on achieving quantum supremacy

In an exclusive interview with MIT Technology Review, Pichai explains why quantum computing could be as important for Google as AI.      by Gideon Lichfield  .... '

Watson Everywhere

Missed this part, a good overview of where IBM plans for Watson to be in the Future.   They will be a part of an AI Future.

Watson Anywhere: The Future
October 21, 2019 | Written by: Rob Thomas

Rob Thomas, General Manager, IBM Data and AI, at the IBM Data and AI Forum in Miami, Fla., Oct. 22, 2019, announcing key updates to Watson Anywhere. 

(Part 3 in a Series) There’s a paradox in the world of AI: While it’s the largest economic opportunity of our lifetime (estimated to contribute $16 trillion to GDP by 2030), enterprise adoption of AI was less than 4% in 2018. A recent Gartner survey said that the 4% in 2018 has now grown to 14% in 2019. But still, that is meager. This is for a variety of reasons: lack of skills, lack of tools, lack of confidence, etc. But the biggest issue is cultural.

For organizations that want to participate in this phase of innovation and wealth creation in technology, the most important thing is a beginner’s mindset; a willingness to try, and an acceptance of failure. Organizations should seek to do 100 AI experiments a year, knowing that more than 50% will fail. Many company cultures are not suited for that. A more typical approach is to rally around one big AI project, committing a lot of people, time and money. I do not advise that approach. AI is about mass experimentation, not one big project implementation. This ain’t ERP.

Fortune favors the bold. I believe that the trial and error all have gone through – and will continue to go through – is worth the positive outcomes. Not just because of the economic opportunity, but the potential to help businesses, consumers, and ultimately, the world in which we live. There will be more experimentation, more failures, more successes. And certainly, many changes to how we live and work. It is up to all of us to ensure that those changes are for the better.

I believe every human being on Earth will interact with Watson in some way – whether it’s accelerating the customer service they receive, augmenting the work they do, improving their retail experiences, providing medical insights to their caregivers, helping them to avoid food scarcity, or even ways that have not been conceived yet. Our ambition has not relaxed. IBM will continue to pioneer AI for all.

Why do I believe this? Because a crucial element for AI to succeed is trust. Companies must be confident that, despite issues of trial and error, they can ultimately trust AI to make meaningful connections and recommendations based on data. So, when it comes to AI, trust will be hugely important in determining which companies succeed and which ones will not. You can say many things about IBM, but I don’t believe anyone thinks IBM is not to be trusted. Our track record as an institution speaks for itself.

Consider our AI client product references. We have more public references in AI than any other company. And, note my choice of words: these are not custom services engagements as references. I’m talking about clients who are using the products that I describe in the first two posts, like Watson OpenScale, Assistant and AutoAI, to name a few. Now, in some instances, do clients hire IBM services (or the services of other systems integrators) to help? Absolutely. But Watson has moved well beyond custom services.

And as more of our clients tell their AI stories, they inspire others to consider, engage and experiment. I’m excited by the scope of adoption, especially across a variety of industries. So far, the most common use cases I see, by industry, are as follows:    .... " 

P&G Sponsors CES Innovation Pitch Contest

Interesting opportunity to see how a major CPG company reacts to a technology pitch.  Considerable detail for participation at the link below.

P&G seeks next big breakthrough via pitch contest
Procter & Gamble Co. intends to again stage a pitch contest at the annual International Consumer Electronics Show in Las Vegas to spotlight innovative technologies or early stage products.

Procter & Gamble Co. intends to again stage a pitch contest at the annual International Consumer Electronics Show in Las Vegas to spotlight innovative technologies or early-stage products.

P&G is usually secretive regarding product development, but the contest reveals the company is interested in all sorts of solutions related to health.

The Cincinnati-based maker of consumer goods such as Align probiotics (NYSE: PG) said the winner will get $10,000, qualify as a finalist in a Techstars accelerator and attend the EY Strategic Growth Forum.

While there’s no guarantee that the winner will get to partner with Procter & Gamble on developing a product, the goal of the second annual P&G Ventures Innovation Challenge is to mine for new ideas.

Created in 2015, P&G Ventures is an early-stage startup studio that creates new brands and businesses by partnering with entrepreneurs, visionaries and startups.

“The Innovation Challenge enables P&G Ventures to discover and work with visionaries who share a mutual dedication to finding solutions that improve the way people live their lives," said Leigh Radford, general manager of P&G Ventures “I truly believe P&G’s next big breakthrough innovation and brand will result from the collaboration of the best of an external partner and the best of P&G.”

Entrepreneurs, inventors and startups are encouraged to submit a product pitch between Oct. 28 and Nov. 24 at ventureschallenge.com. The top three finalists will pitch their products live on the P&G stage on Jan. 8 at the Consumer Electronics Show.  .... 


Legal Implications of Car Hacking

How much regulation is currently out there?  Don't recall signing off on responsibility when buying a car.  Implied?  See the section on legal precedents.  Was unaware of this group within RAND.   We had early connections with them.  Following. 
Who's Responsible When Your Car Gets Hacked?  in Rand.org by Doug Irving

In the future, when cars can drive themselves, grand theft auto might involve a few keystrokes and a well-placed patch of bad computer code. At that point, who will be liable for the damages caused by a hacker with remote control of a 3,000-pound vehicle?

Cars are becoming “fast, heavy artificial intelligences on wheels,” a recent RAND report cautioned—and that means they're becoming vulnerable. Potentially billions of dollars ride on that question of who has the legal responsibility to keep hackers from grabbing the wheel or cutting the brakes.

“These are not likely events, and there are lots of engineers working to make them even less likely,” said James Anderson, the director of the RAND Institute for Civil Justice and a coauthor of the study. “But they're not impossible. They will occur. It's at least worth some serious thought about what the legal consequences will be.”  ... '    

Advances in Quadropod Robotics

Big questions remain, how will we successfully interact with increasingly powerful and complex robot systems?  How do we compare various levels of autonomy to remote controlled robotics?  Inspection is good place to start, having narrow goals.

ANYbotics Introduces Sleek New ANYmal C Quadruped

The latest version of ANYbotics' four-legged robot can do useful real-world inspection tasks   By Evan Ackerman

Quadrupedal robots are making significant advances lately, and just in the past few months we’ve seen Boston Dynamics’ Spot hauling a truck, IIT’s HyQReal pulling a plane, MIT’s MiniCheetah doing backflips, Unitree Robotics’ Laikago towing a van, and Ghost Robotics’ Vision 60 exploring a mine. Robot makers are betting that their four-legged machines will prove useful in a variety of applications in construction, security, delivery, and even at home.

ANYbotics has been working on such applications for years, testing out their ANYmal robot in places where humans typically don’t want to go (like offshore platforms) as well as places where humans really don’t want to go (like sewers), and they have a better idea than most companies what can make quadruped robots successful. ... "

IBM Pushes Back on Quantum Supremacy

 In a largely technical piece in the IBM Research blog.   IBM pushes back on claims of 'quantum supremacy' by Google in some of their recent  tests.  This basically makes the case that context of how the problem and solution are stated is important.  Simulation depends on accurate structure.  The claim remains to be too narrow for claiming supremacy.   Awaiting more details, results, context statements.

On Quantum Supremacy
Written by: Edwin Pednault, John Gunnels, and Jay Gambetta

Categorized: Quantum Computing

Quantum computers are starting to approach the limit of classical simulation and it is important that we continue to benchmark progress and to ask how difficult they are to simulate. This is a fascinating scientific question.

Recent advances in quantum computing have resulted in two 53-qubit processors: one from our group in IBM and a device described by Google in a paper published in the journal Nature. In the paper, it is argued that their device reached “quantum supremacy” and that “a state-of-the-art supercomputer would require approximately 10,000 years to perform the equivalent task.” We argue that an ideal simulation of the same task can be performed on a classical system in 2.5 days and with far greater fidelity. This is in fact a conservative, worst-case estimate, and we expect that with additional refinements the classical cost of the simulation can be further reduced.

Because the original meaning of the term “quantum supremacy,” as proposed by John Preskill in 2012, was to describe the point where quantum computers can do things that classical computers can’t, this threshold has not been met.

This particular notion of “quantum supremacy” is based on executing a random quantum circuit of a size infeasible for simulation with any available classical computer. Specifically, the paper shows a computational experiment over a 53-qubit quantum processor that implements an impressively large two-qubit gate quantum circuit of depth 20, with 430 two-qubit and 1,113 single-qubit gates, and with predicted total fidelity of 0.2%. Their classical simulation estimate of 10,000 years is based on the observation that the RAM memory requirement to store the full state vector in a Schrödinger-type simulation would be prohibitive, and thus one needs to resort to a Schrödinger-Feynman simulation that trades off space for time.

The concept of “quantum supremacy” showcases the resources unique to quantum computers, such as direct access to entanglement and superposition. However, classical computers have resources of their own such as a hierarchy of memories and high-precision computations in hardware, various software assets, and a vast knowledge base of algorithms, and it is important to leverage all such capabilities when comparing quantum to classical.  ... " 

Tele Operating Trucks

Are we expecting this change too quickly?

A big rig just hit 55 mph on a Florida highway without anyone in the cab    By Trevor Mogg 

If you were driving along the Florida Turnpike recently and happened to see a big rig motoring along without anyone behind the wheel, then no, your eyes were not deceiving you.

A system built by San Francisco-based Starsky Robotics powered the driverless truck to speeds of 55 mph along 9.4 miles of public road earlier this month. It’s thought to be the first time a heavy-duty commercial truck has been allowed on a public highway without anyone behind the wheel.

Starsky CEO Stefan Seltz-Axmacher said his company is competing with the likes of Waymo and TuSimple — which still use safety drivers in their autonomous trucks — by taking a different approach to how it incorporates its technology into large vehicles.   ...  "

More on the topic ad its connection jobs.  Autonomous driving mostly remote controlled?

Tuesday, October 22, 2019

(Updated Time) Talk: Sports Highlight Video Creation

Join our talk on the creation of the IBM sports highlights Videos:

Talk by:Stephen Hammer,  Sports CTO, IBM Distinguished Engineer

#CTO & #IBMDistinguishedEngineer Oct 24 10:30-11:00am US Eastern  - "Sports Summary Highlight Videos using AI" -  #CSIGnews #opentechai #ISSIP @KarolynSchalk @ibrahimatlinux @mattganis @odbmsorg @RiyaMRoy @dzwietering @BedangSen

Zoom meeting Link: https://zoom.us/j/7371462221
Zoom Callin: (415) 762-9988 or (646) 568-7788 Meeting id 7371462221
Zoom International Numbers: https://zoom.us/zoomconference
Thu, Oct 24, 10:30am -11:00am US Eastern   @     https://zoom.us/j/7371462221
More Details and recordings will be posted Here : http://cognitive-science.info/community/weekly-update/

Sponsored by our Cognitive Systems Institute and ISSIP (International Society of Service Innovation Professionals) 

Please pass this on to people with interest in video analysis, AI, Sports marketing, etc.




This blog contains topic tags at the end of each post.   I use these to connect topics and posts for my own research and consulting.     These tags do evolve.  Names change, new ideas emerge, some topics split up, old tag names take on new meaning.  Emergent stuff changes.

Recently I had to edit the connections of a number of AI-related tags for specific projects.   I do this on an as-needed basis.  Sometimes on request, sometimes because I just see the need.   So I cannot assure you that especially older tags are appropriate.   Be aware,  ask if you want a clarification.


Machines Collaborating with Humans

Good piece from BAIR: Berkeley Artificial Intelligence Research.  Note the use of games to establish, demonstrate and experiment with collaborative interactions.  YES, there must be a level of appropriate contextual understanding to make collaboration work.   And such collaboration is the best form of assistance.   Descriptions and videos of game play in the below link:

Collaborating with Humans Requires Understanding Them
By Rohin Shah and Micah Carroll        Berkeley

AI agents have learned to play Dota, StarCraft, and Go, by training to beat an automated system that increases in difficulty as the agent gains skill at the game: in vanilla self-play, the AI agent plays games against itself, while in population-based training, each agent must play against a population of other agents, and the entire population learns to play the game.

This technique has a lot going for it. There is a natural curriculum in difficulty: as the agent improves, the task it faces gets harder, which leads to efficient learning. It doesn’t require any manual design of opponents, or handcrafted features of the environment. And most notably, in all of the games above, the resulting agents have beaten human champions.

The technique has also been used in collaborative settings: OpenAI had one public match where each team was composed of three OpenAI Five agents alongside two human experts, and the For The Win (FTW) agents trained to play Quake were paired with both humans and other agents during evaluation. In the Quake case, humans rated the FTW agents as more collaborative than fellow humans in a participant survey.

However, when we dig into the weeds, we can see that this is not a panacea. In the 2.5 minute discussion after the OpenAI Five cooperative game (see 4:33:05 onwards in the video), we can see that some issues did arise1:   ... "