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

Saturday, November 18, 2017

Private Label Grocery

 Is private label grocery about to go to the next level?   by Denise Leathers with discussion .... 

Through a special arrangement, presented here for discussion is a summary of a current article from Frozen & Refrigerated Buyer magazine.

What a difference 12 months makes! Last year at this time, we were resigning ourselves to the fact that private label shares in food may have plateaued — and store brands in the U.S. might never enjoy the kind of success they have in Europe.  .... " 

Now, a new report from Cadent Consulting suggests private label dollar share could swell as much as eight points to 25.7 percent by 2027.   .... " 

Vetting Design with Rapid Experimentation

Reminds me of the idea of develop and test.   A kind of brute force of examination of options.  Always a good way to start thinking about a problem, but does depend on how quickly you can create designs and then test them.

Julie Stanford on vetting designs through rapid experimentation

The O’Reilly Design Podcast: Quickly test ideas like a design thinker.   By Nikki McDonald 

In this week’s Design Podcast, I sit down with Julie Stanford, founder and principal of user experience agency Sliced Bread Design. We talk about how to get in the rapid experimentation mindset, the design thinking process, and how to get started with rapid experimentation at your company. Hint: start small.  .... " 

Downside of Collaboration

Have seen much of this in the enterprise.  From Knowledge@Wharton:
Too Much Togetherness? The Downside of Workplace Collaboration

" ... The mere mention of keeping up with overflowing email, constant meetings, and time-sucking conference calls makes many of us groan and roll our eyes. How did we all get so busy?

A major culprit is the sharp rise in cross-functional collaboration over the past several years. Today, it’s often not enough to just put your head down and work in one department for one boss. Demands can come at you daily from other functional areas of the company — marketing, R&D, finance — sometimes from colleagues you barely know, both within the U.S. and overseas in other time zones. And if you work in an open-plan office, your colleagues will often walk around starting impromptu discussions (when it’s convenient for them, not necessarily you). It can be overwhelming. .... " 

Yet collaboration has long been touted as a key to success. Steve Jobs is credited with saying, “Great things in business are never done by one person, they’re done by a team of people.” Virgin’s Richard Branson recently told Inc., “The fundamental driver of our success at Virgin has, and will always be, our people working together.” Oracle co-founder Larry Ellison has stated, “There is no such thing as a self-made man.” And one can go back further for many old chestnuts about teamwork.  .... " 

Dynamic Programming for the Masses

From ACM Sigsoft a presentation about a favorite topic:  Dynamic Programming:  entitled  'Dynamic Programming for the Masses'.   With links to presentation by  Prof Patrick Madden of SUNY Binghamton.   I will be testing some of the software mentioned, join in if you like, will report on possibilities here as I progress.

Friday, November 17, 2017

AI Now Institute to Research Impact of AI on Society

The Field of AI Research Is About to Get Way Bigger Than Code 
in NextGov.com   by Dave Gershgorn
November 16, 2017

Microsoft Research's Kate Crawford and Open Research at Google founder Meredith Whittaker on Thursday announced the launch of the AI Now Institute, a cross-disciplinary initiative that will research the impact of artificial intelligence (AI) on society. Fields AI Now plans to bring together include data science, law, economics, and sociology. "The amount of money and industrial energy that has been put into accelerating AI code has meant that there hasn't been as much energy put into thinking about social, economic, ethical frameworks for these systems," Crawford says. "We think there's a very urgent need for this to happen faster." AI Now will focus on four core themes, including AI's ramifications for bias and inclusion, labor and automation, rights and liberties, and safety and critical infrastructure. Crawford and Whittaker want their efforts to yield a shared vernacular between AI creators and those who study its effects so societal solutions can be invented.   .... " 

Project Management: Asana and Gmail

Was pointed to Asana as a good system for project management when using Gmail

" ... Asana lets teams track projects from start to finish so everyone knows who is doing what and by which deadline and its current status. Teams that are already using Asana, including Airbnb, General Electric and Nasa can now receive an email, turn it into a task, assign it to a team member, set a due date and track an activity without toggling between email and Asana.  .... " 

More Gmail productivity Add-ons ...

Smart Mirrors

Short piece on the conceptual advances of the 'Smart Mirror'.  Was also a discussion point in our innovation centers.  Could they inform you,  inspire you?

Siri Siri on the wall, is this smart mirror society’s downfall?
An iHome for the modern Snow White     by Thuy Ong  @ThuyOng

Someday mirrors will be smart enough to lie to us, assuring us we’re the fairest of them all. Until then, there’s this thing from iHome. .... " 

Identifying Viral Bots

In O'Reilly,  a considerable look at the problem of bots in social.  Like the use of network diagrams for analytics.   Bringing up the concept of 'information entropy'.   There is so much noise in social this needs to be carefully measured. Note this uses the Polinode graphical network tool, mentioned here previously.

Identifying viral bots and cyborgs in social media
Analyzing tweets and posts around Trump, Russia, and the NFL using information entropy, network analysis, and community detection algorithms.        By Steve Kramer  PhD

" ... In this article, I have demonstrated how it is readily possible to identify social bots and cyborgs on both Twitter and Facebook using information entropy and then to find groups of successful bots using network analysis and community detection. Given the extreme risks of disinformation and propaganda being spread through social media, it is our hope that this approach, along with the work of other researchers, will enable greater transparency and help protect democracy and the authenticity of online discourse. I invite researchers who wish to collaborate on studies of these data sets to request access to become collaborators on our data project hosted on data.world. ... " 

Simulation Tools

Just reexamined the simulation tools that exist today.     Would like to also see more as it applies to process optimization and higher level event based systems.   Found this,  Finite element examples mainly here:

Co-Design Center Develops Next-Gen Simulation Tools      By HPCwire 

The Exascale Computing Project's (ECP) Center for Efficient Exascale Discretizations (CEED) collaborates with application scientists, vendors, and software technology developers to create highly optimized discretization libraries and next-generation mini applications based on advanced high-order finite element methods.

These methods are providing simulation quality without increasing computational time.... " 

Model for Cybersecurity

Defining a New Model for Cybersecurity Trust With Blockchain     By eWeek

The Massachusetts Institute of Technology's (MIT) David Shrier recently detailed the MIT Trust::Data Consortium's efforts to build new systems that rethink data sharing and security.

Shrier said privacy and encryption are important, but a need exists to securely access private data for multiple use cases. "We need a policy around us owning our own personal data and not Facebook and we should also have an audit trail of access," he said. "And we want to be able to share information when we want."

He proposed sharing answers to a query instead of full datasets, and his data collection model would have a query function that communicates with different pieces of code to ensure only authorized queries are accessing the system.

Shrier noted the general trust model can employ blockchain to provide a cryptographically verified distributed system for audits, with computation and query executed in a sandbox to further bolster data security.  .... "

Amazon Go and the Problem of Checkout

Remember using this as a scenario at our innovation centers.  What if a shopper could just walk out, setting up a frictionless way to shop?  Does it bring more shoppers in?   In an era when we tested RFID tagging.   What are the technical issues today?

Amazon's automated convenience stores edge closer to public debut
The company has worked out some of Amazon Go's major technical issues.
By Mallory Locklear, @mallorylocklear

Last year, Amazon opened its first convenience store embedded with its "just walk out technology." Located in Seattle, the Amazon Go store, which lets shoppers walk in, load up on the items they want and walk out without having to pay for the items in a checkout line, has been testing its technology with Amazon employees. Now, as Bloomberg reports, the company has worked through some of the hangups with the technology and is making moves towards opening its store and others to the public. ... " 

Thursday, November 16, 2017

Google and Conversational Apps

Am a long time developer and proponent of intelligent and conversation enabled interactions.

Google supporting Voice and Dialog Based Interactions

Introducing Dialogflow Enterprise Edition, a new way to build voice and text conversational apps
   By Dan Aharon, Product Manager for Cloud AI

From chatbots to IoT devices, conversational apps provide a richer and more natural experience for users. Dialogflow (formerly API.AI) was created for exactly that purpose — to help developers build interfaces that offer engaging, personal interactions.

We’ve seen hundreds of thousands of developers use Dialogflow to create conversational apps for customer service, commerce, productivity, IoT devices and more. Developers have consistently asked us to add enterprise capabilities, which is why today we’re announcing the beta release of Dialogflow Enterprise Edition. The enterprise edition expands on all the benefits of Dialogflow, offering greater flexibility and support to meet the needs of large-scale businesses. In addition, we're also announcing speech integration within Dialogflow, enabling developers to build rich voice-based applications.  
....  "

Also see more about Dialogflow.

Digital Ultimatum

 46 Days Left, P&G's Marc Pritchard on the state of his Digital Ultimatum

By Jack Neff. Published on November 15, 2017. in AdAge

Time's almost up for major digital ad sellers to show that they've improved, Procter & Gamble Chief Brand Officer Marc Pritchard said at Ad Age Next on Wednesday. And that includes big-time players such as Amazon. Hear his take as a year-long ultimatum nears the deadline as well as P&G's latest acquisition in this interview with Jack Neff, following their talk onstage.   .... " 

Caption Generation Tutorial

 Jason Brownlee on caption generation using neural networks, an old problem we addressed.  I like the explanations Jason does.  Get on his instructive solutions mailings.

" ...  Hi, this week we have a gentle intro to neural nets for automatically generating captions for photos ...   Neural network models can automatically describe the contents of a photograph. Discover more in this gentle introduction:

How to Generate Textual Descriptions for Photographs with Deep Learning

Image captioning requires careful preparation of the photograph and text data. Discover how to prepare data for this problem in this tutorial:

How to Prepare a Photo Caption Dataset for Training a Deep Learning Model   .... "

Delivering Ingredients for Allrecipes by Amazon

A long time cook here. Have tested Allrecipes on Echo where it works well, with some reservations about how well the interface works with serial step recipe use.   Vocally at least.

Amazon works with Allrecipes to let you order ingredients for delivery   by Jacob Kastrenakes 

Amazon’s attempted takeover of the grocery store continues in another small way today as the company begins to integrate with the popular cooking site Allrecipes. According to Engadget, Allrecipes will begin allowing its readers to click a button on popular recipes and have the ingredients for that recipe added to their AmazonFresh cart. They’ll then be able to order the products for same- or next-day delivery.  ... " 

Learning from Microsoft's Bots

Like I have mentioned many times, we built consumer facing bots that engaged and delivered expertise.   Saw Microsoft live demo the construction of an Azure based bot lately in a conference.  Good thoughts in this piece:

Lessons in AI from Microsoft’s New Bot

Microsoft and North Highland share insights that go beyond the technology hype. ... Lessons in AI from Microsoft’s New Bot
By Fastco 

" ... Here, Microsoft’s senior channel marketing manager, Chris Kauffman, and Michael Hollar, global lead of AI and cognitive services at North Highland, a global management consultant, share AI insights from the project—a rich case study for any business interested in the technology.   ... "

More on Azure Bots.

November-December Analytics Magazine

From the Informs Organization:

Analytics Magazine: Sneak Preview

The November/December issue of Analytics magazine takes a close look at a wide range of topics, including predictive analytics in the public sector, survey sampling, sales analysis, healthcare analytics and how automated analytics help detect significant business incidents. ... " 

Reconfigurable Camera Systems

Shree Nayar discusses "Cambits: A Reconfigurable Camera System" (cacm.acm.org/magazines/2017/11/222170), a Contributed Article in the November 2017 CACM.

 .... Cambit pieces can be assembled to create a dozen different imaging systems. To celebrate this assortment, Communications has published four different covers, each one featuring a different Cambit configuration. ... 

" ... Here, we present Cambits, a set of physical blocks that can be used to build a variety of cameras with different functionalities. Blocks include sensors, actuators, lenses, optical attachments, and light sources, assembled with magnets without screws or cables. When two blocks are attached, they are connected electrically through spring-loaded pins that carry power, data, and control signals. The host computer always knows the current configuration and automatically provides a menu of imaging functionalities from which the user can choose. Cambits is a scalable system, allowing users to add new blocks and computational photography algorithms to the current set. ... " 

Demonstration video:   https://vimeo.com/236437712

Wednesday, November 15, 2017

More on Wal-Marts Bossa Nova Bots among the Shelves

One of my favorite topics.  How will robots be effectively used in real life environments, contexts?  Note 3D imaging.   And how do you get a count of what is behind?  More at the Bossa Nova Tag.

Walmart Deploys Robots to More Stores
By Patrycja Malinowska - 11/14/2017  in CGT 

Originally published by Path to Purchase Institute .... 

A small test of shelf-scanning robots has gone well enough that Walmart is adding upgraded units to 50 additional stores.

The fully autonomous bots employ 3D imaging to dodge obstacles while roving the aisles to check for stock levels, pricing and misplaced items. The robots pass their shelf data to store employees, who then stock the shelves and fix any errors.

Walmart is hoping applying artificial intelligence and robotics will help it combat stock issues, replenish inventory faster and broaden the assortment available to online shoppers while freeing up employees for other tasks.

The machines are from San Francisco-based Bossa Nova Robotics, stand approximately two feet tall and are equipped with a tower fitted with cameras. They are 50% more productive and can scan shelves significantly more accurately than their human counterparts, Jeremy King, chief technology officer for Walmart U.S. and e-commerce, told Reuters.  .... " 

Private, Portable Deep Learning

More mobile deep machine learning:

New Technology Makes Artificial Intelligence More Private and Portable 
University of Waterloo News  by Matthew Grant

Researchers at the University of Waterloo in Canada say they have developed new deep-learning artificial intelligence (AI) software that could enable other AI technologies to break free of the Internet and cloud computing. The technology is compact enough to fit on computer chips that can be used in mobile devices and industrial robots, which would enable devices to operate independently of the Internet while using AI that performs almost as well as tethered neural networks. The researchers say the technology could lead to much lower data processing and transmission costs, greater privacy, and use in areas where existing technology is impractical due to cost or other factors. The researchers based the new device on evolutionary forces in nature to make the AI more efficient by placing it in a virtual environment, and then progressively and repeatedly depriving it of resources. The deep-learning AI adapts and changes itself to keep functioning each time computational power and memory are removed.  ... " 

Google Aims to put AI Anywhere

Note this says machine learning, not AI.  Which is appropriate.  ML is an important part of AI, but it is not all of it.   From Google: 

Announcing TensorFlow Lite     Posted by the TensorFlow team

Today, we're happy to announce the developer preview of TensorFlow Lite, TensorFlow’s lightweight solution for mobile and embedded devices! TensorFlow has always run on many platforms, from racks of servers to tiny IoT devices, but as the adoption of machine learning models has grown exponentially over the last few years, so has the need to deploy them on mobile and embedded devices. TensorFlow Lite enables low-latency inference of on-device machine learning models.

It is designed from scratch to be:

Lightweight Enables inference of on-device machine learning models with a small binary size and fast initialization/startup

Cross-platform A runtime designed to run on many different platforms, starting with Android and iOS

Fast Optimized for mobile devices, including dramatically improved model loading times, and supporting hardware acceleration  ... '

Marriott Guestroom Lab

We were involved as a corporate customer of hotel space.  And saw many future hotel scenarios to attract the business traveler.  Here is Mariott's take,  collaborating with tech companies:

Marriott Guestroom Lab brings cutting-edge smart tech to your hotel room  By Lulu Chang in DigitalTrend

You’ve taken great pains to ensure that your home is smart, but implementing the same technologies in your home away from home is a bit more challenging. After all, you’re not exactly in a position to tell an international hotel chain how to outfit its rooms. But luckily, you won’t have to, now that Marriott International has teamed up with Samsung and Legrand to debut what they’re calling the “hospitality industry’s Internet of Things hotel room of the future.”

Hotels the world over have of course already begun incorporating high-tech features. There’s Yotel, known for its futuristic rooms geared toward the modern traveler, and of course, the “Strange Hotel” in Japan, which famously features a T-rex receptionist. But Marriott is proving that even established brands can and indeed are looking to bring the 21st century into their guest experience.

It’s all thanks to the IoT Guestroom Lab, which is exploring concepts to “elevate the guest experience, create more efficient hotel room design and construction, and contribute to Marriott’s global sustainability efforts and goals.” The Lab is said to be integrating multiple IoT systems, devices, and applications to speak with one another, and better address the needs of guests and hotel operations staff.  ... " 

Risk Detection is Important

Not often covered well enough, worth thinking about.  Detecting the Risk, working with business partners to understand its implications, is important.  In my observation, not often done well. 

The neglected art of risk detection by  Piotr Kaminski and Jeff Schonert in McKinsey

At the core of risk management is risk detection, an art that can be skillfully improved if banks and regulators accept new analytical methods.

The modern risk-management framework generally relies on the “three lines of defense” scheme, with the businesses, control functions, and audit as the first, second, and third line, respectively. The concept borrows from the language of military strategy, in which intelligence plays a key role. For risk management, intelligence means effective detection: to prevent the bank’s reputation, liquidity, and capital position from being harmed, the lines of defense must detect risks early.

Detection is fundamental in risk management, embedded in its activities and processes. Credit scoring, for example, is a tool for detecting potential borrower-default risk at the application stage, while customer due diligence is designed to identify high-risk customers during the onboarding process, as part of the bank’s know-your-customer (KYC) program. Risk managers are practicing the art of detection when they identify instances of fraud, spot a drifting investment strategy in an asset-management business, monitor their network’s end points to locate cyberintrusions and data theft, or identify potential rogue traders.

Most executives and risk professionals will quickly acknowledge the basic importance of detection. Yet the efficacy of detection—and the levels of “detection risk”—vary widely among risk disciplines and from bank to bank. With poor detection, threats can rise to existential proportions, as some of the world’s largest institutions have learned in recent years. Weak detection capabilities can be costly. Manual controls, for example, are not especially effective and yet they always cost more than automated controls. Poor detection can result in high levels of false positives and the needless diversion of valuable risk resources. .... " 

Machines Writing Music

Delivering a level of creativity is always of interest.

From Mozard to Botzard: when machines write our music
Paris (AFP) - Machines are already taking our jobs, will they soon be writing our music too?

Swiss researchers said Thursday they have developed a computer algorithm that can generate brand-new tunes in different musical genres.

The deep artificial composer, or DAC, "can produce complete melodies, with a beginning and an end, that are completely original," said co-developer Florian Colombo of the EPFL research university in Lausanne, Switzerland.

And the melodies are "quite agreeable to listen to," he told AFP.

The DAC programme uses a form of artificial intelligence known as "deep learning" that works in a similar way to the human brain in memorising experiences and learning from them.

It is a fast-growing field, with more and more possibilities opening up as computers grow stronger and databases larger.

The DAC system is trained to "listen" to existing tunes to learn what works and what does not.

It teaches itself to predict the pitch and duration of every note following another.

Once it is accurate at predicting 50 percent of note pitches and 80 percent of note durations in existing songs, the machine's training is complete.  .... " 

Cognitive Manufacturing


View from the Marketplace
IBM Watson Video Series - Cognitive Manufacturing

By using data and predictive analytics, it’s possible to improve the manufacturing process from the supply chain through to production. ... " 

Tuesday, November 14, 2017

Virtual Beauty in Augmented Reality

We spend some time looking at Virtual Beauty applications.  Was never successful in our day.   What can make it work here?   Is this an example of AR success?   A review.

MAC Virtual Try-on Mirror review
It's crazy how well MAC's Virtual Try-On Mirror can mimic makeup     By Brenda Stolyar in Digitaltrend

Feeling completely lost with makeup? MAC’s Virtual Try-On Mirror lets you try on various looks without making a mess.

Shopping for makeup can be a little overwhelming. While we do have social media and YouTube creators to thank for the influx of beauty gurus and makeup tutorials, it’s common to think you won’t be able to pull off stellar and dramatic looks. But what if all you needed was augmented reality (AR) to give you that extra boost of confidence? MAC Cosmetics’ new beauty mirror gives users the opportunity to try on different makeup looks in real-time, with the help of ModiFace’s AR technology ... "

Fashion AI by Alibaba

Examined the use of this kind of in store technology to engage. Personalization and privacy were key aspects.   But we also discovered that the environment of such interactions had to be carefully controlled.

Alibaba’s AI Fashion Consultant Helps Achieve Record-Setting Sales

AI will blur the line between online and offline retail.

by Yiting Sun in Technology Review

In the third floor of a shopping mall in the heart of Shanghai last week, Xiaolan He, a woman in her 50s, took an olive-green down jacket to a fitting room. To her surprise, she found a screen about the size of a large poster on the wall. It recognized the item of clothing in her hands through a tiny sensor embedded in the garment, and showed several options for matching items that she could flip through like a photo album. The screen, and the system that powers it, make up FashionAI—which essentially became He’s personal stylist.   .... "

Smart Pills Approved

An old idea we talked in the innovation centers:

Smart Pills Are Headed to American Patients
The FDA has approved America’s first digital pill. The tablets, called Abilify MyCite, are an antipsychotic which can be used to treat schizophrenia and some cases of bipolar disorder. But unlike a regular pill, they contain a small ingestible sensor … " 

Google Home Adds Intercom

Been using a similar function in Amazon Alexa for some time, its a popular function in the Smart Home.   Especially if you have a larger space.  Only makes sense if you have have more than one device. Google has added some clever and humorous functions to the intercom.    Always thought Amazon was ahead in making such assistant functions more 'human' and engaging, nice to see Google Home Advancing here.   More below.

Stop yelling across the house — use your Google Home as an Intercom instead    By Lulu Chang  in DigitalTrends ... "

Lowes and Macy's Want to Help with the Smart Home

More retailers join into the outfitting of the smart home.

Lowe’s and Macy’s join rivals chasing smart home opportunity

by Tom Ryan   and Expert Comments in Retalwire

"Retailers have a finite period of time to convince customers that they are the destination for smart home devices."   -  Joan Treistman

Move over Best Buy and Amazon.com, Lowe’s and Macy’s want to also help consumers transform their homes with the Internet of Things (IoT).

Last week, Lowe’s announced it is opening 70 in-store shops dedicated to helping consumers discover and buy smart home devices.

The rollout follows the successful pilot of SmartSpot shops in three Lowe’s locations last fall operated by b8ta. The collaboration with the software retailer will continue. Described as “retail-as-a-service,” b8ta’s model relies on manufacturers paying a monthly subscription fee to be in the store.... " 

Media Companies vs Amazon

More detailed thoughts at the link:

Should media companies fear Amazon?  by Ted Chamberlin in Gartner

In weighing both sides, I see Amazon as being a clear and present danger to the established media establishment.  Amazon the supply chain in media and entertainment; they might not opt to build every link ( players, mobile devices,) but are keenly aware of how cord-cutters and cord-nevers will be choose OTT video over linear broadcast.  Here is a great graphic provided by my colleagues Monica Zlotogorski and Fernando Elizalde on how viewer habits are changing ... " 

Monday, November 13, 2017

Limits of Machine Learning

Good piece.  In particular was struck by the term 'brittleness', which we encountered many times in out exploration of enterprise AI.   A solution could be useful in a given context, but if the context changed even slightly, and most do change over time, do we detect it, what is the risk, and how to we prepare for any changes?    The challenge in modern machine learning is no different.  Every project we worked had an element of this. 

The Limits of Machine Learning – Is your ML Solution Viable?

By Ahmed Fattah  Executive Analytics Architect at IBM Cognitive Solutions


Machine Learning (ML) is an extremely powerful technology that is likely to transform business and society. In some cases, ML is the perfect tool for a given task in others it could be an overkill or utterly inappropriate. As with any technology, ML has limits. Lack of understanding of these limits is likely to result in inappropriate use, undesirable outcomes and eventually a repeat disenchantment with whole of AI. The key tools for understanding and taking account of these limits and for developing effective and viable ML solutions are the proven disciplines of Software Engineering and Architecture.


When I was writing my Machine Learning (ML) thesis I came across this quote:
“The statement ‘God created man in His own image’ is recursive.”
I was extremely enchanted, inspired and empowered to develop my intelligent ML algorithm. I believe that the above quote encapsulates the fascination we hold of ML and the deep-seated desire to create intelligent machines. At the same time, the quote conjures some fear and uneasiness. These ambiguous perspectives fuel hype, inflate expectations and polarize people into zealots and alarmists. As an AI practitioner, I am, of course, very enthusiastic about the recent rise of AI and ML and confident that these technologies will transform business and society. However, with my experience in Software Engineering and Architecture, I am aware of the danger of inflated expectations and infatuation with new over-hyped technology.

As with many past over-hyped technologies, there is always a danger that arises from misunderstanding their essence and limits. That is why we should always have a healthy dose of doubt when anything is promoted as a silver bullet. Although ML is not the first technology to be over-hyped, ML has unique characteristics that exacerbate this danger. ..... " 

Multiple Speaker Recognition

Assistants can now recognize individual voice patterns.  Have trained my own that way, usually by repeating phrases.   And that categorization is important to understand context. But this is usually in a forced, single person situation.  More broadly, the 'cocktail party' problem is harder,  beyond just a voice print, also using the context of the words spoken.  Takes it beyond controlled environment dialog.  Good challenge.

An AI has learned how to pick a single voice out of a crowd   By Richard Gray

Devices like Amazon’s Echo and Google Home can usually deal with requests from a lone person, but like us they struggle in situations such as a noisy cocktail party, where several people are speaking at once.

Now an AI that is able to separate the voices of multiple speakers in real time promises to give automatic speech recognition a big boost, and could soon find its way into an elevator near you.

The technology, developed by researchers at the Mitsubishi Electric Research Laboratory in Cambridge, Massachusetts, was demonstrated in public for the first time at this month’s Combined Exhibition of Advanced Technologies show in Tokyo.

It uses a machine learning technique the team calls “deep clustering” to identifies unique features in the “voiceprint” of multiple speakers. It then groups the distinct features from each speaker’s voice together, allowing it to disentangle multiple voices and then reconstruct what each person was saying. “It was trained using 100 English speakers, but it can separate voices even if a speaker is Japanese,” says Niels Meinke, a spokesperson for Mitsubishi Electric. .... "

Retailwire on AI's Role in Customer Engagement

From Retailwire:

Why artificial intelligence is becoming instrumental for unifying in-store and online experiences ... After decades of intense development, AI is coming into its own in retailing.

AI is a technology poised to exercise tremendous influence on the way businesses operate, inspiring retailers to reimagine what their enterprises can accomplish.

Download your copy of this new digital whitepaper from RetailWire to learn how retailers are putting AI to use for natural language processing, predictive analytics, building shopper profiles based on their behavior, pattern recognition and more.

See this m•Paper for insights on:

AI’s growing influence in retail
What we’ve learned from digital commerce
AI in the store — a data explosion
AI and the mobile experience
How AI is driving omnichannel  ....  " 

Sunday, November 12, 2017

Supply Chain Digest Numbers

Supply Chain Digest publishes sort term news items, based on statistics.  here is an example.  More at the link, but still thin on actual details.   Worth following. 

Supply Chain by the Numbers for Week of Nov. 9, 2017
On-Line Grocer Ocado Automates DC for Success; Deal with the Devil on Chinese Aircraft? Carrier Now Offering Minimum Weekly Driver Pay; Net US Store Closings are Very Much on the Rise ... " 

Process Mining

Process mining     From Wikipedia, the free encyclopedia ....

Process mining is a process management technique that allows for the analysis of business processes based on event logs. During process mining, specialized data-mining algorithms are applied to event log datasets in order to identify trends, patterns and details contained in event logs recorded by an information system. Process mining aims to improve process efficiency and understanding of processes.[1] Process mining is also known as Automated Business Process Discovery (ABPD).[2] ... " 

Hype of Virtual Medicine

Good points are made here,  in particular are systems like smartphones, effective at addicting us otherwise, useful for health habits?   WSJ article:

The Hype of Virtual Medicine
High-tech health care hasn’t proved effective at changing patients’ bad habits.... 

Dolby Labs Sensing Biometrics

Not a new thing, we worked with several startups in this domain, still being implemented as 'Neuromarketing' in consumer in context measurement.   With lots of variability in how it is being done, from simply measuring galvanic skin response, to full FMRI scans.  Have seen and experimented with the data from both.  Both for marketing focus and for reaction to product in ad or shelf context.  See tags below.  First I had heard of Dolby in this space.

Next Level: Inside Dolby's Experiment to Watch People While they are Watching Movies
Measuring emotional responses could be the next big thing in entertainment   by Lauren Goode in TheVerge.

" .. Led by neurophysiologist and Dolby chief scientist Poppy Crum, the company has been attaching biosensors to willing subjects and plopping them down on a couch to settle in for an entertainment session. Using EEG caps, heart rate monitors, galvanic skin response sensors, and thermal imaging Flir cameras, the scientists can observe the biophysical and emotional responses that humans are experiencing via media. They’re trying to figure out what kind of videos and sounds make people’s hearts race, what makes their skin flush, and what makes them cognitively engaged, aroused, or maybe even... bored. ... " 

Kirio, An Installed Learning Smarthome

Builders installing sensors to make homes adaptive.

“Kirio creates smart homes from scratch, and they learn as you live
By Jenny McGrath in Digital Trends

With its stainless-steel appliances, open-plan style, and large windows, you could easily mistake the tri-level home currently for sale at 2621 Franklin Ave East for any other pricey Seattle property. Though you might walk into a dining room with gentle yellow lighting and soft music playing, you might not automatically peg it for a smart home. The only thing that might give you a clue is the white, rounded hexagon placed on the wall next to a Nest thermostat. The Kirio runs the home, though, and will soon do so for about 300 more houses in the Seattle area.

We first saw a demo  of the integrated smart home system this summer. Back then it took place in Kirio’s office. It was a little different to see it all come together inside a three-story condo that costs nearly $1.2 million. Seattle is one of the country’s hottest housing markets, but Kirio thinks its little white hub (though the team doesn’t like it to call it that) can still make properties stand out. Instead of a DIY system homeowners buy after moving into a place, builders install Kirio along with a bunch of sensors and other devices during construction. It controls everything in the home — “wired and wireless,” CEO Rob Green told Digital Trends during a tour of the home. “It’s a real, live smart house.” .... ' 

Saturday, November 11, 2017

Resume Assistant for Linkedin

Always thought there was good opportunity here.  More still exists by linking capabilities to needs, and in meetings to tasks to be done.   And how about linking to dialog in email streams?

"Resume Assistant” uses LinkedIn’s data to make Word a better résumé builder
One of the first fruits of the LinkedIn purchase should make an unpopular task easier.  By Peter Bright

 Writing and updating your résumé is a task that few of us enjoy. Microsoft is hoping to make it a little less painful with a new feature coming to Word called Resume Assistant.

Resume Assistant will detect that you're writing a résumé and offer insights and suggestions culled from LinkedIn. LinkedIn is a vast repository of both résumés and job openings and lets you see how other people describe their skillsets and which skills employers are looking for.

The feature will also show job openings that are suitable for your résumé directly within Word, putting résumé writers directly in contact with recruiters.

Matching up job applicants with recruiters is LinkedIn's biggest revenue generator, and integrating Word into this business was one of the immediate priorities that Microsoft CEO Satya Nadella described when explaining the purchase. In addition to bolstering that recruiting business, the integration is monetized more directly, too, as Resume Assistant has some connections to paid services. Specifically, it can promote online training through LinkedIn Learning to help shore up any gaps. If you want more help than the automated suggestions can offer, LinkedIn can connect you to professional résumé writing services through its freelance marketplace. ..... " 

Twitter Posts

My Twitter stream https://twitter.com/franzd  contains links to many of the posts here.   I also add related posts from twitter and other social sources a similar range of topics.   Follow along there and add your own references.

Guide to Chatbots

A guide and survey of chatbots.  Looks good.  Will follow.

Beginners Guide to Chatbots  Posted by William Vorhies 

Summary:  This is the first in a series about Chatbots.  In this first installment we cover the basics including their brief technological history, uses, basic design choices, and where deep learning comes into play.  In subsequent articles we’ll describe in more detail about how they are actually programmed and best practice dos and don’ts.  ... " 

Six Analytics Books

KD Nuggets books of interest to the data scientist.  Nice selection.  Mostly in the moderately technical domain.  Some of these are more background books that are good reads rather than 'how to'.

Gartner on Digital Retail Strategies

Most Retail Digital Strategies Not Yet Harvesting Results   by Robert Hetu   

The retail CIO agenda for 2018 reflects the existential threat to many retailers, caused by digitally enabled consumers, new entrants and new business models in the market. Gartner’s annual CIO Survey gathered data from 3,160 CIO respondents in 98 countries and across major industries, representing $13.0 trillion in revenue and public-sector budgets and $277 billion in IT spending. This report focuses on the answers from 169 retail CIO respondents and compares them with the total sample. .... " 

Friday, November 10, 2017

Life After Driving

Imagining a self-driving world.  Life After Driving

A very broad look at self driving vehicles, the near term and unexpected social and technical consequences.

In  the NY Times Mag.   Nicely done.

Who Wants an Amazon Key?

Still intrigued by the Amazon Key as a means to open a unique channel to the consumer.   But who wants it, who fears it, and what do they expect to get out of it?  Can it deliver?  Recode looks at some numbers.

Most Amazon Prime subscribers say they don’t want to buy the Amazon Key that lets delivery people into their homes

But 5 percent of Prime subscribers would definitely buy an Amazon Key.
 By Rani Molla @Ranimolla  .... "

Google Maps Show Checkout Line Length

Adding line lengths to maps of stores.  Saw a similar approach done at a Kroger, but just for their store.  The data will be crowd sourced, which can lead to incorrect numbers.   With expert discussion:

Has Google solved the problem of long lines at grocery checkouts?   by Matthew Stern

Grocers have tried plenty of tech solutions over the years to shorten the time customers spend in line. But the latest wait-shortening technology is coming not from an individual grocer, but Google’s crowdsourced data.

Google Maps is implementing functionality that uses anonymous crowdsourced location data to let users see estimated wait times at grocery stores, according to Thrillist. Google will roll the technology out before Thanksgiving. The tool for determining the wait at grocery stores is the second of two data-driven crowd size-gauging tools. The first, which launched Tuesday, allows users to determine what the average wait time will be for a table at a given restaurant.

Grocer strategies for limiting customer frustration with lines have taken many forms. Some have tried distraction, installing digital signage to make the time in line pass more quickly. Others, like Hy-Vee, have tried data-driven “traffic lights” that gauge the business of each line and signal each customer to the fastest checkout.  ... " 

Chinese Economics

Been looking at the 'One Belt, One Road Initiative'.  Earlier we had looked at modeling retail demand in Chinese markets using agent simulation modeling, as it predicted future supply chains.    See china tag and 'Agent Models' below.  This Wharton view is of interest. Continuing to examine.

Will President Xi Jinping Let Markets Decide China’s Future?

China’s twice-a-decade Communist Party congress wrapped up on October 26 with party head and President Xi Jinping claiming a place alongside revolutionary leader Mao Tse-tung in the pantheon of modern Chinese heroes. In addressing the congress, Xi focused on contradictions the party faces in steering China toward a brighter future, much as Mao highlighted “class contradictions” decades earlier.

But the party gathering did little to clarify how the leadership plans to handle the challenges simmering within the world’s second-largest economy, and especially within its chaotic financial system.

Franklin Allen, an emeritus Wharton professor of finance, and a professor of finance and economics at Imperial College in London, questions Xi’s appetite for painful reforms that are needed to rebalance China’s economy away from its longtime heavy reliance on construction investment and exports. “Xi has established tremendous political power. Now the question is how much of that he will turn into economic reforms,” Allen says. For now, Xi seems to prefer using his “One Belt, One Road Initiative,” to expand Chinese supply chains and market access, as well as its geopolitical influence, across the developing world. In the meantime, on the domestic front, “I would not be surprised if there was slow and gradual reform.” .... "  

Baidu Launches Conversational OS

Ultimately it will be system that can natural converse with us that will be AI.  Not just systems that can recognize language, but systems that can carry on an ongoing, intelligent, useful dialog based on changing context.  Still not done by the assistants out there.  Work by Baidu below is of interest, is this beyond just natural language processing?   See also work by Facebook.

Baidu launches new operating system dedicated to conversational AI   By Duncan Riley in SiliconAngle

Chinese search giant Baidu Inc. today officially launched a new operating system dedicated to advancing conversational artificial intelligence capabilities, as the company aims to take on similar efforts from competitors such as Google Inc.

Called the “Duer OS Prometheus Project,” the operating system is a conversation-based AI platform that is already providing conversational support to 10 major domains and more than 100 subdomains in China. Launched in beta at the beginning of the year, the code base is claimed to be the top choice already for third-party hardware manufacturers in China in areas such as TV settop boxes, storytelling machines and smart speakers that are seeking Mandarin language voice recognition support.   .... " 

See more on Baidu and related assistants relating Dialog based AI in this blog.  Updated frequently.

Thursday, November 09, 2017

State of Machine Intelligence

Well done piece, with useful charting out of the state of intelligence.  Below just a brief excerpt:

The current state of machine intelligence 3.0
Watching the appeal and applications of machine intelligence expand.
By Shivon Zilis, James Cham  in O'Reilly 

" .... What's the biggest change in the last year? We are getting inbound inquiries from a different mix of people. For v1.0, we heard almost exclusively from founders and academics. Then came a healthy mix of investors, both private and public. Now overwhelmingly we have heard from existing companies trying to figure out how to transform their businesses using machine intelligence.

For the first time, a “one stop shop” of the machine intelligence stack is coming into view—even if it’s a year or two off from being neatly formalized. The maturing of that stack might explain why more established companies are more focused on building legitimate machine intelligence capabilities. Anyone who has their wits about them is still going to be making initial build-and-buy decisions, so we figured an early attempt at laying out these technologies is better than no attempt ..... " 

AlphaX Decision Sciences

Brought to my attention,  Advanced Analytics for Oil Companies.

About AlphaX: 

AlphaX Decision Sciences, LLC is incorporated in Delaware with offices in Houston, San Jose and Dubai. The Company’s ten employees are dedicated to bringing enterprise-wide advanced analytics software products and services to our customers, who include independent oil companies (IOCs), national oil companies (NOCs) and oilfield service companies (OFS). We choose to keep private the names of our customers in light of the competitive advantage that our work provides. Our technology infrastructure inherently addresses security and data privacy concerns and our organization is designed to deploy projects within a structured, yet agile process. Our data science and software development teams hold advanced degrees in mathematics, physics and electrical engineering while our subject matter experts are industry veterans. Whether you are looking for a point solution in analytics or looking for how to design your organization to address the competitive advantages for the future we can assist you in unlocking the value of your data and learning to monetize it by building machine learning or deep learning intellectual property. .... " 

Internet of Risky Things

Risk increases as things exist in external environments, as does the IOT.

The Internet of Risky Things - What to do about IOT Sceurity?  By Peter Liss in Linkedin   ... Expert Business Developer in IoT, Telco and Technology

In Part 3 of my IoT articles, the focus is on IoT specific security. What makes IoT different to conventional I.T security? What is the extra cost? What needs to be considered that might not already be implemented in your existing I.T security?

There is general agreement that IoT Security is a big worry, but there are different interpretations of what to do. Is the biggest worry with the platform, types of connectivity, communication protocols, applications, chipsets, devices, server, firmware, or sensors? Every security expert I ask has different answers, depending on whether they are a software of network or hardware person. As the saying goes, “give someone a nail and they will use a hammer ...  " 

Wednesday, November 08, 2017

How to Store Information in Your Clothes

Quite remarkable idea, ready to test this out before buying more storage devices.

How to Store Information in Your Clothes Invisibly, Without Electronics
By University of Washington 

A new type of smart fabric developed at the University of Washington could pave the way for jackets that store invisible passcodes and open the door to your apartment or office.

The UW computer scientists have created fabrics and fashion accessories that can store data — from security codes to identification tags — without needing any on-board electronics or sensors.

As described in "Data Storage and Interaction Using Magnetized Fabric," presented at UIST 2017, the ACM Symposium on User Interface Software and Technology, the researchers leveraged previously unexplored magnetic properties of off-the-shelf conductive thread. The data can be read using an instrument embedded in existing smartphones to enable navigation apps.

"This is a completely electronic-free design, which means you can iron the smart fabric or put it in the washer and dryer," says senior author Shyam Gollakota, associate professor in the Paul G. Allen School of Computer Science & Engineering. "You can think of the fabric as a hard disk — you're actually doing this data storage on the clothes you're wearing."     ..... " 

Data Science: Challenges and Directions

Late to this, but very useful.  Full text at link.  Have always been confused as to the very definition of data science, as opposed say to Analytics.  science or optimizing a process task?  This article scopes it well.

Data Science: Challenges and Directions
By Longbing Cao 

Communications of the ACM, Vol. 60 No. 8, Pages 59-68

While data science has emerged as an ambitious new scientific field, related debates and discussions have sought to address why science in general needs data science and what even makes data science a science. However, few such discussions concern the intrinsic complexities and intelligence in data science problems and the gaps in and opportunities for data science research. Following a comprehensive literature review,5,6,10,11,12,15,18 I offer a number of observations concerning big data and the data science debate. For example, discussion has covered not only data-related disciplines and domains like statistics, computing, and informatics but traditionally less data-related fields and areas like social science and business management as well. 

Data science has thus emerged as a new inter- and cross-disciplinary field. Although many publications are available, most (likely over 95%) concern existing concepts and topics in statistics, data mining, machine learning, and broad data analytics. This limited view demonstrates how data science has emerged from existing core disciplines, particularly statistics, computing, and informatics. The abuse, misuse, and overuse of the term "data science" is ubiquitous, contributing to the hype, and myths and pitfalls are common.4 While specific challenges have been covered,13,16 few scholars have addressed the low-level complexities and problematic nature of data science or contributed deep insight about the intrinsic challenges, directions, and opportunities of data science as an emerging field .... " 

Is Big Data Done?

Depends upon what you mean by Big.  Has to be enough, correct, accessible, actionable, process usable,  stable, etc.     We all manage by data, have done that for a long time, but now we have more much of it,  so better to do that carefully.  Good cautions though.  Considerable discussion at the link.

Big Data is done, put a fork in it
By Rick Ferguson, CLMP  in Retailwire  

Through a special arrangement, presented here for discussion is a summary of a current article from The Wise Marketer, a website and newsletter serving the global loyalty industry.

About a decade ago, the phrase “data is the new oil” swept the globe as the twin corporate power centers of IT and marketing realized their companies had more data than they knew what to do with. “Drinking from the fire hose” became a metaphor for the struggles of extracting actionable insights from data.

About five years ago, the usual suspects in IT consulting and cloud-based analytics began to trumpet the phrase “Big Data” to sell into companies hoping to extract that resource.

Now, as Slate’s Will Oremus points out, the phrase Big Data has become passé — in part because we just call it data now, and in part because the rush to rely solely on data for business decision-making has often revealed the limitations of data-based decisions.

Our over-confidence in these tools often hinders our ability to see the forest for the trees. We often fall victim to what data scientist Shane Brennan calls the “Ten Fallacies of Data Science.” For a variety of reasons, the data may be inaccessible, indecipherable, outdated, lacking enough granularity for analysis or prohibitively expensive to get or have processed, Mr. Brennan contends. Often, a marketer’s lack of understanding or pre-conceived notions leads to poor decisions.

“Garbage in/garbage out” results in marketing campaigns or loyalty offers that are ineffective or even harmful to your business.  .... " 

Distill: For Machine Learning Research

Brought to my attention: Distill

Distill:  Machine Learning Research

Should Be Clear, Dynamic and Vivid.
Distill Is Here to Help.

A modern medium for presenting research

The web is a powerful medium to share new ways of thinking. Over the last few years we’ve seen many imaginative examples of such work. But traditional academic publishing remains focused on the PDF, which prevents this sort of communication.  .... " 

Publishing in the Distill Research Journal

Why Publish in Distill?

Flexibility Distill encourages you to go beyond traditional academic forms. The goal is to best communicate science and serve the reader.

Recognition Distill articles are peer reviewed and appear in Google Scholar. Distill is also registered with the Library of Congress and CrossRef.   
ISSN: 2476-0757    DOI: http://doi.org/10.23915/distill
Neutrality Distill provides a neutral platform for multiple authors to jointly publish. This is in contrast to a personal website, where outside contributors may get less credit.
Impact Distill articles are read by tens of thousands of people.
Journal Scope

Distill is an academic journal in the area of Machine Learning. The distinguishing trait of a Distill article is outstanding communication and a dedication to human understanding. Distill articles often, but not always, use interactive media.

One good test for whether your article is a fit for Distill is whether your collaborators and you are willing to put in whatever time is necessary to write and illustrate an outstanding article. In our experience, this often takes 100+ hours. Typically, we expect this to mean that there is at least one collaborator who is very enthusiastic about explaining things well.    .... " 

Feature Visualization

In the Google Blog, they provide an overview and pointer to deeper information.  An image rich and informative article.

Feature Visualization
Posted by Christopher Olah, Research Scientist, Google Brain Team and Alex Mordvintsev, Research Scientist, Google Research

Have you ever wondered what goes on inside neural networks? Feature visualization is a powerful tool for digging into neural networks and seeing how they work.

Our new article, published in Distill, does a deep exploration of feature visualization, introducing a few new tricks along the way!

Building on our work in DeepDream, and lots of work by others since, we are able to visualize what every neuron a strong vision model (GoogLeNet [1]) detects. Over the course of multiple layers, it gradually builds up abstractions: first it detects edges, then it uses those edges to detect textures, the textures to detect patterns, and the patterns to detect parts of objects….   " 

" .... There is a growing sense that neural networks need to be interpretable to humans. The field of neural network interpretability has formed in response to these concerns. As it matures, two major threads of research have begun to coalesce: feature visualization and attribution. ... " 

Alexa as the End of Smartphones?

Had never thought of it quite this way. overall a big challenge.  Mobile is still convenience.  Also not enough about the attentiveness of such systems.  Voice/conversation in intelligent dialog is still crude.

For Amazon, The Future Of Alexa Is About The End Of The Smartphone Era
With touch-screen devices like the Echo Show and Echo Spot, Amazon is trying to upend a decade of smartphones, notifications, and apps.    By Jared Newman in FastCOMpany

Picture a hypothetical future in which Amazon’s Alexa renders your smartphone unnecessary.

You tap the “good morning” button on your bedside Echo Spot, and it displays the day’s forecast as the blinds open and the bedroom lights fade on. Downstairs, the Echo Show in the kitchen displays a reminder about your doctor’s appointment at noon, and alerts you to a single urgent email: You’ll have to join a conference call as soon as you get into the office.

In the car, you press the “commute” button on the Alexa-enabled dashboard, which queues up your favorite music while steering you around a traffic jam, and after work, Alexa relays another reminder to your car, telling you to grab milk on the way back. At home, you swipe through a list of restaurants on the Echo Show to make a reservation for the weekend, check in on your folks via video chat, and finally sit down to relax. With one more buttonpress, the next episode of your favorite show starts playing on the Fire TV in the living room.

Far-fetched as it seems, this is the kind of scenario Amazon wants to create as its Alexa assistant migrates from smart speakers to touch screens. Instead of making users wade through a sea of apps to get things done, the goal is to create a new kind of computing that’s simpler and less distracting. And while voice will be the foundation, there may be plenty of times where you use Alexa without speaking at all.   

“The tenet that we believe in is that voice is a simplifier,” says Miriam Daniel, Amazon’s head of product management for Alexa. “And once you design voice actions, then go back to a touch action, you’ve actually removed a lot of friction because you’ve started at a different baseline.” .... ' 

Tuesday, November 07, 2017

Valens Offers a Single Copper Cable

Quite interesting solution that extends the simple .... for the rolling supercomputer.

Valens Automotive Offers Carmakers One Cable To Rule Them All   By Philip E. Ross in IEEE Spectrum

As our cars turn into rolling supercomputers, it’s easy to forget one of their emerging electronic needs: managing massive data flows within the car. In a self-driving car, sensors should throw off some 4 terabytes a day.

“Connectivity is one of the missing elements,” says Micha Risling, head of the automotive business at Valens, a fabless semiconductor company in Israel. “On one hand you have the Nvidias of this world—the companies that offer processing solutions—and, on the other hand, there are the companies that offer sensors, cameras, even displays. We provide the connectivity.”  

And because Valen’s method involves just a twisted pair of copper wires, a single cable can replace several. That’s an improvement—even over the futuristic alternative, which is to use fiberoptic cable. Carmakers would rather put off that wrenching change and just extend the Age of Copper. .... '

Sony Advanced Robot Aibo Dog

Saw the early version of the Aibo dog demonstrated in Japan.    Tried to get a demo for our space, but it never happened.   Japan was and is interested in eldercare: 'friendly' robot solutions.  Reminds me a bit of the Kuri robot, which aims to be a friendly family member, with some assistant components.  The Alexa assistant has been described as friendly,  with hints of being a family member.   Is this a component that will sell home robotics?  Will mere friendliness sell? Note the many sensors included.

Sony Unleashes New Aibo Robot Dog    by John Boyd

Eighteen years after unveiling its original Aibo robot dog, Sony is now reviving the iconic product using advanced mechatronics and AI to create a cuter, smarter, and more lifelike version. The new Aibo is crammed with ultracompact actuators, two cameras, touch sensors, microphones, and a 64-bit quad-core CPU. According to a Sony executive, the robot uses AI to create "a bond with its owners that can grow over time." Sony will start selling Aibo early next year. ... " 

Cisco Adds a Spark Conferencing Assistant

Said to be a result of Cisco's acquisition of MindMeld earlier this year.  Makes sense, now will it add a level of attentiveness to a conference and do supporting searches and create to-do lists?

With Spark Assistant, Cisco adds voice-activated A.I. to videoconferencing   By Matthew Finnegan 

Cisco envisions a phased rollout of Spark Assistant, a new service designed to end some of the pain out of setting up office meetings. ... " 

AI as the New Electrcity

Interesting thoughts on AI,  Still beware of hyped expectations.  Brains in Alexa or Siri?   Still awaiting that.  In Knowledge@Wharton. 

Why AI is the new Electricity

Just as electricity transformed the way industries functioned in the past century, artificial intelligence — the science of programming cognitive abilities into machines — has the power to substantially change society in the next 100 years. AI is being harnessed to enable such things as home robots, robo-taxis and mental health chatbots to make you feel better. A startup is developing robots with AI that brings them closer to human level intelligence. Already, AI has been embedding itself in daily life — such as powering the brains of digital assistants Siri and Alexa. It lets consumers shop and search online more accurately and efficiently, among other tasks that people take for granted.

“AI is the new electricity,” said Andrew Ng, co-founder of Coursera and an adjunct Stanford professor who founded the Google Brain Deep Learning Project, in a keynote speech at the AI Frontiers conference that was held this past weekend in Silicon Valley. “About 100 years ago, electricity transformed every major industry. AI has advanced to the point where it has the power to transform” every major sector in coming years. And even though there’s a perception that AI was a fairly new development, it has actually been around for decades, he said. But it is taking off now because of the ability to scale data and computation. .... "

On Memory Video Holograms

As the Verge article linked to beow  suggests, the term Hologram is being used differently and non-technically here, as kind of 3D video that can be later accessed using VR tech for interaction.  Read the whole article to understand what is being done in this space.  Is this a popular future channel for consumer remembering?

Are holograms the future of how we capture memories?
Holographic time capsules     by Lauren Goode  @LaurenGoode 

" ... holograms, a term that’s used (and sometimes overused) to describe volumetric video. 8i was looking to build up its repository of holograms, and a mom-and-baby clip seemed like a good idea for relatable content.

Scott’s first hologram taping led to her crying in a VR headset, and became what the company’s chief executive called a “watershed moment” for the then-fledgling company. ... " 

See also about the company mentioned:

8i: Mix realities with holograms:

Easily record and experience photorealistic human holograms with true volume and depth for virtual, augmented and mixed reality ... " 

Mentors, Mentees, Assistants

Perhaps obvious, but reasonable statement of the needs.  In the HBR.
Can a virtual assistant be a mentee?  We examined.  In the short term, no, the information exchange can be subtle, but longer term it can present structure for basic knowledge. 

 What Mentors Wish Their Mentees Knew
Vineet Chopra, MD, Sanjay Saint, MD

" ... The mentor-mentee relationship is a tango between a more senior person and a junior one. Just as in dance, coordination and orchestration between parties is necessary for grace and success. And while we and others have written about what makes the ideal mentor, comparatively less has attention has been given to the other partner. This gap is unfortunate because, like mentorship, menteeship requires specific behaviors — without which the mentee’s success may be threatened. In this article, we outline five habits of ideal mentees and provide anecdotes and views from our combined years of academic experience. While we focus on the relationship in academic medicine, the takeaways apply to most any field.  ... " 

Hinton and Capsule Networks

We followed neural networks from their earliest days, but where they failed they did not have the data to consistently work under changing context, and at high enough speed, say for analyzing real-time video.  Geoffrey Hinton's work brought a new approach forward.  Now he addresses some of the lingering problems with these methods:  speed, error rates and the need for huge amounts of data. Tech papers pointed to in the link below.

Google's AI Wizard Takes a New Twist on Neural Networks in Wired, by Tom Simonite.

" ... But Hinton now belittles the technology he helped bring to the world. “I think the way we’re doing computer vision is just wrong,” he says. “It works better than anything else at present but that doesn’t mean it’s right.”

In its place, Hinton has unveiled another “old” idea that might transform how computers see—and reshape AI. That’s important because computer vision is crucial to ideas such as self-driving cars, and having software that plays doctor.

Late last week, Hinton released two research papers that he says prove out an idea he’s been mulling for almost 40 years. “It’s made a lot of intuitive sense to me for a very long time, it just hasn’t worked well,” Hinton says. “We’ve finally got something that works well.”

Hinton’s new approach, known as capsule networks, is a twist on neural networks intended to make machines better able to understand the world through images or video. In one of the papers posted last week, Hinton’s capsule networks matched the accuracy of the best previous techniques on a standard test of how well software can learn to recognize handwritten digits.

In the second, capsule networks almost halved the best previous error rate on a test that challenges software to recognize toys such as trucks and cars from different angles. Hinton has been working on his new technique with colleagues Sara Sabour and Nicholas Frosst at Google’s Toronto office.

Capsule networks aim to remedy a weakness of today’s machine-learning systems that limits their effectiveness. Image-recognition software in use today by Google and others needs a large number of example photos to learn to reliably recognize objects in all kinds of situations. That’s because the software isn’t very good at generalizing what it learns to new scenarios, for example understanding that an object is the same when seen from a new viewpoint.   ...." 

Monday, November 06, 2017

Flying Cow Drone Cell Service

A drone that temporarily flies at a few hundred feed to provide services.   Our own look at commercial drone services showed this could be done.  40 mile square radius is impressive.  Expect more of this kind of solution to be coming commercially.

AT&T's 'Flying COW' drone provides cell service to Puerto Rico
The mini-helicopter can provide voice, data and text services in a 40-square-mile radius.  By Rob LeFebvre, @roblef ... "   in Engadget

Building AI Applications to Solve Analytics

Interesting view of how different companies are addressing the complex issue of how AI solutions are built and delivered.     Getting more important by the day. 

Building AI That Can Build AI  
The New York Times,  by Cade Metz  (May require signup and registration) 
November 5, 2017

Google is concentrating on developing artificial intelligence (AI) that can partly relieve humans from constructing AI systems that many think represent the future of the technology industry. Firms are inventing all varieties of tools that will make it easier for any operation to build its own AI software, including products such as image- and speech-recognition services and online chatbots. Scientists such as Google's Jeff Dean (who shared the 2012 ACM Prize in Computing with Sanjay Ghemawat) believe if more people and companies are working on AI it will drive their own research, while companies such as Google and Microsoft envision revenue-generating opportunities; all are selling cloud computing services that can help other businesses and developers build AI. Dean thinks Google's AutoML project, which seeks to automate neural-network construction, will help firms build AI systems even if they lack expertise. Google is building algorithms that analyze the development of other algorithms, learning successful and unsuccessful methods so they eventually learn to build more effective machine learning.  .... " 

Developing Enterprise AI Apps

Now been involved in a number of such enterprise efforts, nicely put, can use more examples and details.   See the use cases linked to for more.

Developing Successful AI Apps for the Enterprise

The IBM team encourages developers to ask tough questions, be patient, and be ready to fail gracefully.   By Nicole Tache November 6, 2017,  Network exchange in O'Reilly

Read about use cases and practical tips for bringing AI to the enterprise in a chapter excerpt from “Getting Started with Artificial Intelligence.” Download the free chapter “Natural Language Processing.”

In this episode of the O’Reilly Media Podcast, I sat down with Josh Zheng and Tom Markiewicz, developer advocates for IBM Watson. We discussed how natural language processing (NLP) APIs, and chatbots in particular, represent just one of the ways AI is augmenting humans and boosting productivity in enterprises today.

In order to apply AI to the enterprise, Zheng and Markiewicz explain, developers first need to understand the importance of sourcing and cleaning the organization’s data, much of which is coming in unstructured formats like email, customer support chats, and PDF documents. This can be “unglamorous” work, but it’s also critical to building a successful NLP app, or chatbot.

From there, Zheng and Markiewicz offer some practical tips for developers looking to build chatbots: to have context awareness, to fail gracefully, and to have patience—building a successful chatbot can take time. .... " 

Where AI has Gone Wrong

Most of the examples here are well known.  Many of them are less the AI going wrong but how they have been inserted into the underlying business process, sometimes driven by poor expectations.  This is not that informative for the practitoner, but since this is published in a well known investment site, it is how the capabilities are being presented, and thus good caution to all involved.

6 Scary Stories of AI Gone Wrong
Technology can have a mind of its own.     by Chris Neiger (TMFNewsie) in TheFool

Artificial intelligence is poised to become a nearly $60 billion market by 2025 and you'd be hard-pressed to find a major technology company that's not investing in AI technology right now. 

But the rapid pace at which AI is growing means that sometimes the technology is being tested in the wild before it's been properly vetted in the lab. And, in other cases, even carefully crafted AI systems tend to act in ways that their developers never anticipated.

Here are just a few examples of how artificial intelligence has gone wrong, sometimes comically -- and other times times a little more eerily.  ..... " 

Ocado Robots Filling Grocery Orders

With video and detail.    Impressive move forward..  Have now watched grocery pick up orders being composed by workers in a standard Kroger store.   Not a fast or even very logical process. 

How Many Robots Does It Take to Fill a Grocery Order?
It once took online grocer Ocado two hours to put together a box of 50 food items. Now machines can do it in five minutes. .... 

By Jing Cao in Bloomberg

Images Can be Constructed to Fool Deep Learning

A troublesome result,  especially where there is incentive to do so.

Fooling Neural Networks in the Physical World with 3D Adversarial Objects

We’ve developed an approach to generate 3D adversarial objects that reliably fool neural networks in the real world, no matter how the objects are looked at. ....  " 

More in MotherBoard.

Sunday, November 05, 2017

Building Deep Learning Applications

Intriguing piece I am reviewing, showing the continued complexity of the field, from a practice perspective.

Nuts and Bolts of Building Deep Learning Applications: Ng @ NIPS2016 via DSC   This article was written by Tomasz Malisiewicz.

You might go to a cutting-edge machine learning research conference like NIPS hoping to find some mathematical insight that will help you take your deep learning system's performance to the next level. Unfortunately, as Andrew Ng reiterated to a live crowd of 1,000+ attendees this past Monday, there is no secret AI equation that will let you escape your machine learning woes. All you need is some rigor, and much of what Ng covered is his remarkable NIPS 2016 presentation titled "The Nuts and Bolts of Building Applications using Deep Learning" is not rocket science. Today we'll dissect the lecture and Ng's key takeaways. Let's begin.  ....  "

Cautions About Betting on Platforms

 Most of us are today betting on a platform, the Internet, so this seems odd advice.  But we are also betting on some specifics of the internet, access, architecture and more.    Good cautions.

3 Reasons Why You Shouldn't Bet Your Business On A Platform
The basic laws of business have not changed ...   By Greg Satell in Inc.

" ... However, the basic laws of business have not changed. You still need to create, deliver and capture value in order to compete. In some cases, the best way to do that will be a platform, but in the vast majority of cases it will not be..... " 

Looker for Self Service Analytics

Brought to my attention by a client:  Looker
Claims for democratized and secure, real time data analytics.

It’s a delicate balance between making data available for people to visualize themselves and being the bottleneck in a never-ending river of data definitions. Looker helps data analysts walk the line and make data visualization and exploration easy for everyone.

Learn How Looker Helps You:

•   Merge and maintain accurate data models
•   Address unique problems with custom metrics
•   Blend any data that is logically related
•   Turn data agnostics into data believers  .... " 

Bixby Assistant Expanded?

I think that Samsung's Bixby will be important to follow as an assistant. Saw a prototype of this long ago.  On a visit to an appliance store recently I saw a Samsung advanced refrigerator that included a sign implying that the Bixby assistant was coming.  Bixby has only previously been seen live on Smartphones.   Samsung has a large share of Appliances in the US, so its integration could prove of interest, with implications for the smart home.  More interesting to me than on the phone.  In SiliconAngle, Upcoming announcements:

What to expect from Samsung Bixby 2.0 when it’s unveiled Oct. 18   By Collen Kriel

Retails and Click and Collect

Include further information and discussion.

Are retailers getting comfy with click & collect?   by Nikki Baird in Retailwire.

Through a special arrangement, what follows is a summary of an article from Retail Paradox, RSR Research’s weekly analysis on emerging issues facing retailers, presented here for discussion.

During a panel at the CONNECT Mobile CX Summit in Philadelphia, panelists from Jersey Mike’s, Kohl’s and Brookstone (formerly Advance Auto Parts) all talked about unintended consequences from their early efforts at click & collect.

Issues include dealing with customers showing up at the wrong store because they didn’t realize where they’d actually placed the order. Another challenge, according to the panelists, is making sure a promotion to get people used to ordering online for in-store pickup actually has the scale in stores to support possible over-sized success.

Not surprisingly, the importance of mobile was also talked up at the conference. Once a click & collect order is in flight, most consumers rely on mobile for further communications and updates from the retailer. At the store, employee-facing mobile is becoming more important as retailers are finding manual processes breaking down in the face of all that volume.  ....