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

Monday, July 31, 2017

Neural Compute Stick

Would like to understand what 'plug and play' means here.  Acceleration likely.  Intriguing thing at this level.  wrote about Movidius previously. They were acquired by Intel in 2016, a natural idea.   More details at the link.

This little USB stick is designed to make AI plug-and-play
The Neural Compute Stick from Movidius makes it easy to add a machine vision processor to any device    by James Vincent     @jjvincent   in TheVerge.

Amazon Tries Mobile Flash Sales

Amazon has no intention of stopping their innovative sales ideas. I am imagining the Echos, soon to be announcing sales on your very street.   Will this link Amazon yet closer to the consumer? Called Treasure Trucks.  Reminds me of their purchased Woot flash sales site, but mobile.   More expert discussion at the link in Retailwire.

Has Amazon created another high-draw shopping model with its Treasure Trucks?   by Laura Davis-Taylor

This past Saturday, Amazon.com stealthily launched a flash sale in Atlanta, Chicago, Dallas, Houston and Los Angeles using its only once-before-seen Treasure Trucks.

Touted as “an ice cream truck for grown-ups,” the engagement model, which was tested in Seattle before being rolled out to new markets this weekend, is unique and intriguing. The truck drives around town with a single “amazing deal” and, when it’s going to be in your area, you get a text on your phone. If interested, you click to buy on your app, pick one of the trucks as your pickup point and get over there — fast.

The launch video for the service eludes to items such as new releases and electronics, steaks and seafood, outdoor gear, toys and more. All you do is sign up and wait for the truck to arrive at a city near you  .... " 

Learning by Thinking

Can we make machines work this way?   In a sense we do. we train and re-train.  And test against reality.  But is the semantic language or architecture correct to make this work efficiently?  Not yet.

In the Edge: 

Learning By Thinking
A Conversation With Tania Lombrozo 

Sometimes you think you understand something, and when you try to explain it to somebody else, you realize that maybe you gained some new insight that you didn't have before. Maybe you realize you didn't understand it as well as you thought you did. What I think is interesting about this process is that it’s a process of learning by thinking. When you're explaining to yourself or to somebody else without them providing feedback, insofar as you gain new insight or understanding, it isn't driven by that new information that they've provided. In some way, you've rearranged what was already in your head in order to get new insight. .... " 

Is Web Scraping Illegal?

A common practice under fire again.   If something is clearly publicly delivered, is it open to turn into another form called data?  A challenge to data democratization?

LinkedIn: It’s illegal to scrape our website without permission
A legal scholar calls LinkedIn's position “hugely problematic.”    By Tomothy B. Lee  in Ars Technica... 

Conversational Commerce

Another look at conversational commerce.   Though beyond what we think of as messaging.   Including memory of previous interactions and leveraging what is known about the correspondents. Positioned as a service with value.  Examples at the link:

Why your business should pay attention to Conversational Commerce  in Unified Inbox: 

In an article published by Customer Think (http://bit.ly/2w3vhGy), Amrita Bhattacharyya writes: “Conversational Commerce refers to using natural language within a messenger application (Messenger, WhatsApp, WeChat, and others) or using voice assistants (Siri, Amazon Alexa, and others) to interact with a business for an inquiry, purchase, or customer service.”

Conversational Commerce enables a person to, quite literally, message their refrigerator! For example, you would text your refrigerator “Do we have milk?” Using a combination of Artificial Intelligence (AI) and Natural Language Processing (NLP), your refrigerator would text you back “Yes, but at your current rate of consumption, it will be gone tomorrow. Would you like me to place an order with Peapod?”

This technology is not science fiction from Star Trek, it’s actually available today!

Here are a few more interesting scenarios:   .... " 

Sunday, July 30, 2017

P&G Cuts $100 Million in Digital Ads

P&G Cuts More Than $100 Million in ‘Largely Ineffective’ Digital Ads
Consumer product giant steers clear of ‘bot’ traffic and objectionable content ...

Almost all of the consumer product giant’s advertising cuts in the period came from digital, finance chief Jon Moeller said on its earnings call Thursday. The company targeted ads that could wind up on sites with fake traffic from software known as “bots,” or those with objectionable content.

“What it reflected was a choice to cut spending from a digital standpoint where it was ineffective, where either we were serving bots as opposed to human beings or where the placement of ads was not facilitating the equity of our brands,” he said. ... " 

(Full access may require subscription)

Alexa Needs to be Much Smarter

Have now been testing the Google Home, Amazon Echo, Cortana and Siri since their inception.  All of these devices and services do some clever things, but none are 'intelligent' in even a broad sense of the word. The intelligence they might be said to have is due to the data and external services they are connected to.  The conversations they can have with us are still rudimentary.  And we have to change our speech to ensure the advisor understands our request.

I have yet to see even an individual skill that is remarkable, in a cognitive reasoning sense, beyond what we already expect of computing devices:   Fast Search, calculate, connect, simple advice in a narrow domain.  The winner of this space will take the advisor beyond this.

Amazon's race to make Alexa smarter
By Leo Kelion, Technology desk editor of the BBC

Amazon's range of smart speakers and their artificial intelligence assistant Alexa have proved to be a huge sales hit.

But the product is still a shadow of what the man in charge - Dave Limp - and indeed their owners, hope it will become.

"We have thousands of engineers inside Amazon adding to [its] capability every day and then another tens of thousands of developers adding to the skills," he tells the BBC.

"The thing I am sure of is that this time next year she will be significantly more intelligent than she is now, and that sometime in the future we will hit our goal of reinventing the Star Trek computer." ...'

Index to all my writings on Virtual assistants.

Fewer Assistants in China

Attended a meeting this week at Columbia featuring proposed uses of virtual assistants,  excellent ideas, many by Chinese students.  Impressive.   So I see some evidence to the contrary.

This Is Why China Hasn't Jumped on the Smart Speaker Bandwagon   in Bloomberg 

- Two million units seen shipping versus 35 million in U.S.
- Mobile devices seen as key way to deliver AI-based services ... "

Knowing the Data

I like the idea of 'thinking with data',  not just solving a particular problem but knowing important process data and manipulating it for ongoing value.  Knowing the data as well as you know the analytic methods.

thinking with data with "Modern Data Science with R"
One of the biggest challenges educators face is how to teach statistical thinking integrated with data and computing skills to allow our students to fluidly think with data.  Contemporary data science requires a tight integration of knowledge from statistics, computer science, mathematics, and a domain of application. For example, how can one model high earnings as a function of other features that might be available for a customer? How do the results of a decision tree compare to a logistic regression model? How does one assess whether the underlying assumptions of a chosen model are appropriate?  How are the results interpreted and communicated?  .... "

Good example included.  Includes some free downloads of chapters from associated book.

See also 
SAS and R    Examples of tasks replicated in SAS and R ....

Saturday, July 29, 2017

Mobile Neural Networks

This kind of work is going on elsewhere, like at Google, to try to put the intelligence and learning on the 'edge' of a network.

MIT Research Could Make Neural Networks Portable
By Ryan Whitwam  

Computers aren’t just getting faster these days–  they’re getting smarter. That’s thanks largely to artificial neural networks, which consist of multiple nodes of processing designed to interpret and understand data more like a biological brain. Neural networks are behind modern software features like voice recognition, computer vision, and even beating humans at Go. However, neural networks need a lot of power, so they usually run in the cloud. Having a neural network locally on your mobile device could be hugely beneficial, and it may be possible thanks to research from a team at MIT. ... " 

AI is not Smart Enough Yet

Gets back to the definition of AI.  Sets of algorithms,  machine learning,  rules are still not AI (Artificial Intelligence) , they are new kinds of data processing that will augment people, just like all computing does.   It used to be that computing vastly increased speed and data scope of decisions.   Now decisions can have elements of cognition we could not do before, but as the article gives examples of,  they are not fool-proof.   Thus the use by some of the term 'Augmented Intelligence'.   I like this term, but it is seen by many, especially the investors, as a cop out.

Artificial intelligence is not as smart as you (or Elon Musk) think
Posted Jul 25, 2017 by Ron Miller (@ron_miller)   .... " 

Towards an Intelligent Automation Continuum

I like the broad idea.  We seem to be doing very well these days in picking off specific narrow problem domains, but less so in constructing general solutions to things like conversations, which are ultimately the way our intelligence is delivered and absorbed.   I still don't see the 'continuum' suggested, just very clever quanta, intermixed with increasing amounts of poorly organized data.   But its proposal is a start.

Revisiting the Intelligent Automation Continuum
April 21, 2017 | Tom Reuner  ....

New in Analytics, AI and Automation

Via McKinsey:

Innovations in digitization, analytics, artificial intelligence, and automation are creating performance and productivity opportunities for business and the economy, even as they reshape employment and the future of work.  .... " 

Friday, July 28, 2017

Road ahead for Connected Vehicles

The Road Ahead for Connected Vehicles  In Knowledge@Wharton

Mass-market driverless cars are farther off in the future than media hype would suggest, and big automakers will maintain major strategic advantages. ... "

Prescriptive Business Models

Another great piece from O'Reilly.  Have seen many business models done with simple spread sheets, and even complex analytics built on top of these spreadsheets.   And these are thus,  prescriptive models.   Nice to see some coded up in this way, with source code. But do keep it as simple as possible.

A lesson in prescriptive modeling
Simulate new business models and practices with open source code.
By Jerry Overton July 25, 2017  .... " 

Google Glass Based Augmedix

 Google Glass based, brought to my attention.  Good example of Glass usage.

Augmedix

Rehumanizing Healthcare  .... 

Augmedix gives providers more time to focus on patient care
Regardless of which electronic healthcare record (EHR), providers are spending more time in front of the computer and less time face-to-face with patients. This hurts practices through lost revenue, 
reduced provider and patient satisfaction, and decreased charting accuracy.  Augmedix eliminates the clinic’s biggest pain points

Augmedix gives providers a team of real-time, quality-controlled, and customized remote scribes, accessed through Glass. Our productivity platform takes care of many major administrative tasks. ... " 

Amazon Healthcare Tech

Amazon has skunkworks underway to look at healthcare technologies.   In SiliconAngle:

" ... The lab is said to be working on projects including pushing and pulling data from legacy electronic medical record systems. Presumably such a service could tie into Amazon’s ever-growing Amazon Web Services business.

Other projects reportedly include a telemedicine platform that would facilitate virtual consultations and tap into stored medical data, and health applications for existing Amazon hardware such as the Echo and Dash Wand. The report added that it’s not clear whether Amazon is looking to build new dedicated health devices, but it’s not impossible it might do so in the future.  ... " 

Intel has its Ginger Assistant

Somewhat belatedly brought up, but Intel reported buying Ginger in 2014.   No reports of what has been done with it, but it was mentioned in a Intel sponsored Columbia meeting today.   Following up.

Intel Pays Up To $30M For A Personal Assistant Platform From Ginger Software   Posted May 9, 2014 by Ingrid Lunden

Apple has Siri, and now Intel has Ginger. The chipmaker has made one more acquisition to bolster its advanced computing and artificial intelligence holdings: it has purchased selected assets, and hired talent, from Israel’s Ginger Software in the area of natural language processing tools and applications. Those assets include a platform for third parties to create customised personal assistants, for a price believed to be up to $30 million.

Ginger Software, backed by investors like Li Ka-Shing’s Horizons Ventures, will continue to operate as an independent business focusing on its remaining business: intelligent grammar and spell checking software. This is not the only change afoot at Ginger: the company recently saw its chairperson, Soffer Teeni, leave to head up Facebook Israel.  .... " 

An Overview of State of Virtual Assistants

In PC Mag looking at what you can do with virtual assistants.   A reasonable overview, for those who have not taken a deeper dive.

Thursday, July 27, 2017

How Much Training Data is Required?

In Machine Learning Mastery.  Good place to start, but there is rarely a precise answer.  More is usually good.  A look at one of the most common questions to be asked of the analyst.  Or worse, not asked at all.  In today's world of having huge amounts of data, it still does not mean we have the data needed to ask the question at hand.   Or don't have the meta data to support the conclusion.  Or don't understand how the context in which the data was gathered changes the solution.

Alexa Defies and Defines the Moment of Truth

From Wayfind, thought provoking.  No longer a moment of truth, but a moment of speech?  But isn't the cylinder now in control of the conversation?   Sure Brands will still influence the conversation by their value equity, but control is slipping.

Alexa is Brand Agnostic
What happens if brands no longer influence the moment of truth?     By Lee Peterson

Alexa, order a 12-pack of toilet paper, a case of bottled water, and a bag of tortilla chips.

Welcome to the new moment of truth. There is no shelf. There is no cart. There is no register. There isn’t even a laptop or mobile phone. There is only a talking cylinder on a kitchen table. Artificial intelligence has enabled every household to become a store shelf. And in this new, demand-based moment of truth, something is often being left unsaid: Charmin, Dasani, and Tostitos. Or Scotts, Evian, and Santitas. Or even Cottonelle, Fiji, and Mission.   .... "   (read the whole thing) .... 

P&G Cuts Digital Spending, Sales Rise

So digital advertising not directly tied to sales here.   Stock is up.   Requires sign-in/subscription:

In Adage:    P&G Slashes Digital Ads by $140M Over Brand Safety. Sales Rise Anyway   Net earnings rose 15%, with nearly half the gain from those digital media cuts. P&G might not rush to restore the spending. ....  " 

More detail inYahoo Finance.

IOT vs IIOT

Overview that is useful ... with examples of applications.

IOT vs IIOT   Do you know the difference?
Not all Devices are IoT or IIoT

Posted by Daniel Ehrenreich  in IOT Central

Business opportunities created by Internet of Things (IoT) and the Industrial IoT (IIoT) are among the most debated topics, as these are designed to function in a broad range of consumer and industrial applications.

Manufacturers of IoT components believe in this new trend, but many of them still not understand the essence of the IoT concept. In reality, not every controlled device is an IoT nor IIoT.

The IoT/IIoT concept is a communication-based eco-system in which control devices, CCTV cameras and industrial sensors communicate via the Internet with cloud-based computer systems and data sources, and the result of this process is displayed on a computer screen, smartphone or used for optimal activation of a process. Through an IoT/IIoT ecosystem you may boost productivity and achieve unique benefits. Examples of IoT/IIoT include applications such as; remote operation of home appliances, medical devices, check on availability of a product in a store, warnings of unusual conditions and malfunctions and more.  .... " 

Development Workflows for Data Science

Nicely done free Book/PDF on the topic.  My experience is most people don't consider workflow at all, so even an overview of a resource like this is worthwhile.  Consider setting up a standard for methods in your company.  I like the fact that several options were considered.   Requires minimal registration.

" ... Development Workflows for Data Scientists

GitHub partnered with O’Reilly Media to examine how data science and analytics teams at several data-driven organizations are improving the way they define, enforce, and automate development workflows—including:

Defining team structure and roles
Asking interesting questions
Examining previous work
Collecting, exploring, and modeling data
Testing, documenting, and deploying code to production
Communicating the results

This illuminating report shows how, even though the pace of change is rapid and the desire for the knowledge and insight from data is ever growing, the dual disciplines of software engineering and data science are up for the task. .... " 

Wayfind Research on Amazon and History

Research pieces in Wayfind regarding Amazon and Customer Experience

Current Issue
- Alexa is Brand Agnostic
- If it’s the Amazon Century, how did we get here?
- Amazon’s Place in History
- The Customer You’ve Never Met
- Will the Amazon & Whole Foods Merger Work?
- What If your only options were Amazon, Walmart and Third Wave stores?

One extract:

Amazon’s Place in History    By The WD Strategy & Insights Team

Amazon’s growth, and let’s face it, domination, has not been an accident. Yet it has come at the cost of some…retail salespeople, store closures and even Walmart. This year marks the 3rd Amazon Prime Day, let’s take a look at how Amazon has grown and what it means for the future of the industry.

By some estimates, there is as much as 7 billion square feet of retail space in America. There are now 70 Amazon fulfillment centers,5 some as large as one million square feet, in operation. There are no reliable figures on the square footage of each location, but if an Amazon fulfillment center averages about a million square feet, a rough estimate would be 70 million square feet. .... " 

Completion Obsession

Recall making some very informal observations in this space.    It also appears to be a particular obsession with some people.  A completion obsession?   Application for work, for retail, advertising? Irrational, or some comfort/nesting compulsion?

People Have an Irrational Need to Complete 'Sets' of Things  
People are irrationally motivated to complete arbitrary sets of tasks, donations, or purchases—and organizations can take advantage of that, according to new research by Kate Barasz, Leslie John, Elizabeth Keenan, and Michael Norton.  In HBS Working Knowledge

by Carmen Nobel

Wednesday, July 26, 2017

Model Ensembling to Reduce Error

Technical but nicely done piece, with code and intuitive arguments as to value.  Via O'Reilly:

An introduction to model ensembling by Jovan Sardinha

Model ensembling represents a family of techniques that help reduce generalization error in machine learning tasks. In this article, I will share some ways that ensembling has been employed and some basic intuition on why it works.  .... " 

Open Innovation at Fujitsu

Via Ideaconnection.com

Open Innovation Gateway
is dedicated to activating innovative practices — faster

Open Innovation Gateway uses a collaborative approach to innovation that mirrors Fujitsu’s brand promise of “shaping tomorrow with you.” Our promise - about shaping the future of business and society - invokes our way of doing business, building on long-term relationships with our customers, partners and society as a whole. ... 

At Open Innovation Gateway we look for more than technologies. We look for partnerships, relationships. When we brought in Oblong's amazing Mezzanine technology that began a relationship between the two companies that continues to grow and deepen.   Video. ... 

Science of Brute Force

A technical point, but with important implications.   Analysts look for Algorithms, or simplified statements of what will be solutions in specific contexts.   These predict.   Algorithms can be a set of math or logical statements, or can be a universal proof of some truth we can apply.  It turns out that some algorithms are best addressed by brute force, or by exploring their context space, by iterating through many, many potential solutions.  This is called 'brute force', and is covered in the most recent issue of the Journal of the CACM, linked to below.   Also see this overview Vimeo presentation.   Technical.

The Science of Brute Force
Many relevant search problems, from artificial intelligence to combinatorics, explore large search spaces to determine the presence or absence of a certain object. These problems are hard due to combinatorial explosion, and have traditionally been called infeasible. The brute-force method, which at least implicitly explores all possibilities, is a general approach to systematically search through such spaces. ... " 

Brute force has long been regarded as suitable only for simple problems. This has changed in the last two decades, due to the progress in Satisfiability (SAT) solving, which by adding brute reason renders brute force into a powerful approach to deal with many problems easily and automatically. Search spaces with far more possibilities than the number of particles in the universe may be completely explored. ... " 

Automated Machine Learning

Excellent piece on this topic. Am in the process of preparing a talk to a Columbia University Group on exactly this topic. Useful detail at the link. I think it is inevitable we will see highly automated processes of these types, for professionals and even end users.  Here a good list of work underway and implications for professionals doing the coding.

Automated Machine Learning for Professionals   Posted by William Vorhies  
Summary:  There are a variety of new Automated Machine Learning (AML) platforms emerging that led us recently to ask if we’d be automated and unemployed any time soon.  In this article we’ll cover the “Professional AML tools”.  They require that you be fluent in R or Python which means that Citizen Data Scientists won’t be using them.  They also significantly enhance productivity and reduce the redundant and tedious work that’s part of model building. ... " 

Another China Smart Speaker

But available only in China, and in Mandarin.  Still the considerable number and low costs are interesting.  Following Alibaba's entry.

Xiaomi’s take on the Amazon Echo smart speaker costs less than $50   by Jon Russell  in TechCrunch

Hot on the heels of reports that Facebook is developing its own take on Amazon Echo, China’s Xiaomi has joined the tech company masses by jumping into the increasingly-crowded smart speaker space.

The Mi AI Speaker is Xiaomi’s first take at rivaling the Echo, which has already inspired a product from Alibaba in China and counts offerings from Google and Apple among its competitors.

Building on the voice-controlled speaker that Xiaomi shipped in December, the new device is powered by artificial intelligence, the company said, which has just been added the Xiaomi’s MIUI operating system, a variant of Android. The speaker can be used as a control for Xiaomi products and also smart products from Xiaomi’s one-hundred-plus partners. Xiaomi touted its content that is available for the speaker, which includes music, audio books, kids stories and radio.

In terms of audio itself, the device uses a setup of six microphones for 360 degree sound broadcast. ... " 

See my list of all the virtual assistants and smart speakers covered.  Continues to be updated frequently.

The Flaw of AI Implementation

Disagree with the 'fatal' part,  but it is a major problem.  Been talking this since the last AI summer. Its best to model or at least sketch out the business process first, before you start, to make sure you are addressing the right thing, with the right data.  Get the owner/users of the process involved early, and repeatedly during development.

The Fatal Flaw of AI Implementation     by Jeanne Ross  from MIT Sloan.

There is no question that artificial intelligence (AI) is presenting huge opportunities for companies to automate business processes. However, as you prepare to insert machine learning applications into your business processes, I’d recommend that you not fantasize about how a computer that can win at Go or poker can surely help you win in the marketplace. A better reference point will be your experience implementing your enterprise resource planning (ERP) or another enterprise system. Yes, effective ERP implementations enhanced the competitiveness of many companies, but a greater number of companies found the experience more of a nightmare. The promised opportunity never came to fruition.

Why am I raining on the AI parade? Because, as with enterprise systems, AI inserted into businesses drives value by improving processes through automation. But eventually, the outputs of most automated processes require people to do something. As most managers have learned the hard way, computers can process data just fine, but that processing isn’t worth much if people are feeding them bad data in the first place or don’t know what to do with information or analysis once it’s provided.

Tuesday, July 25, 2017

Myth of the Machine Learning Black Box

Was unable to get to see this, but of interest.  Most all machine learning projects need to be updated,  maintained and reevaluated.   Maybe they are not black boxes .... but are effectively so, because the decision makers have to do a considerable amount of work to figure them out.

The Myth of the Machine Learning Black Box
Added by Tim Matteson on June 21, 2017 

Critics describe machine learning as a "black box," where data goes in and a prediction comes out, without visibility into how the prediction was derived. This lack of transparency makes it difficult to evaluate and update predictive models as conditions change or new sources of data become available. But today's machine learning systems are not black boxes, allowing data scientists and business professionals alike to understand how a model makes its predictions.

In this Data Science Central webinar, DataRobot will discuss how today's automated machine learning systems provide the information and visualizations that deliver deep insights that break out of the black box.

Speaker: Greg Michaelson, Director of DataRobot Labs -- DataRobot

Hosted by: Bill Vorhies, Editorial Director -- Data Science Central

Why AI Can't Write this Article

We jousted with Roger Schank at the time of the first AI eruption, very well known at that time, and I continue to track his work.  Recognizing and understanding stories are a key part of our intelligence. See the tag below.     In the HBR on the state of AI today.  Thoughtful piece.

Why AI Can't Write this Article (Yet),  by Walter Frick

Roger Schank, a reearcher and former professor, once proposed a novel goal for artificial intelligence: A computer should be able to watch West Side Story and recognize the plot of Romeo and Juliet. Schank and his students believed that stories are central to intelligence, reasoning, and meaning. By Schank’s measure, today’s AI isn’t intelligent at all. ... "

Gamifying the E-Commerce Experience

Something we explored, with only a bit of success.    Not specifically endorsing this vendor, Iterable, but the report is interesting.  Requires sign-in.

Game On
What E-Commerce Can Learn From the Top 100 Mobile Games
Gamification is becoming a key strategy for e-commerce companies, from developing loyalty programs to creating an urgency to buy.

Hit all the right psychological triggers of your customers by acting like a gaming company and building a captivating brand.

Iterable Report
Here's what you'll find inside this guide:

Strategies to align your messaging to user behavior
Tips to energize the shopping experience from inbox to checkout
3 tried-and-true steps to keep customers coming back    .... " 

HoloLens 2 Adding AI Processor

Augmented realities will need better edge intelligence to provide value.   By John Brandon, Computerworld.  Some interesting applications are described.

Why Microsoft is adding an AI co-processor to HoloLens 2

With more processing power for AI, the HoloLens 2 could help us better interpret data and make better business decisions ... ' 

" ... Microsoft knows this, and that’s why a recent announcement about adding an AI co-processor to the HoloLens 2 caught my eye. The chip will help with tagging data in the real world, which sounds just as complex and compute intensive as you can imagine. Also, this processing will work much faster when the chip runs locally instead of in the cloud ...  "

Voicelabs Ads

Advertising and Alexa Skills.    


Amazon says it updated its policy to make sure users have the best experience possible.  by Mariella Moon, @mariella_moon in Engadget: 

When VoiceLabs debuted its interactive ads for the Amazon Echo, it had backing from big names like ESPN and Wendy's. For a while, it gave developers using the service called "Sponsored Messages" a way to make money from their Alexa Skills. Unfortunately, even its powerful pals can't do anything for it now that Amazon has updated its advertising policy to ban ads using a voice that's similar to Alexa and those that imitate Alexa interactions. Since those are exactly what the company's interactive ads do, it has no choice but to stop offering Sponsored Messages altogether.

If you've never heard a Sponsored Message before, here's how VoiceLabs describes its product:

"For example, a consumer would be asked if they were going to watch the NBA finals game airing that day. If the user said yes, it would let them know to tune in for the 6pm start time. If not, the Sponsored Message would let them which channel to tune into at 10pm for all the highlights."  ..... ' 

Retail Analytics and Store Viz

We also worked on this kind of analytics.  ...   Good points made, its all about context.

Retail Analytics and Store Visualization   by Gary Angel  In CustomerThink:

Making sense of behavioral data is always a challenge. Suppose I tell you that a shopper visited your store, spent fourteen minutes, lingered twice and had a single Associate interaction. That’s a lot of data, but in most respects, it’s deeply uninformative. What did the shopper care about? What were they interested in? What did they pass by?  What worked? What didn’t? You can’t even begin to formulate answers to those questions based on the data I described. 

I know, because I spent years trying and largely failing to do interesting analytics with metrics like these in the digital world. What’s missing from these metrics is the context. If you don’t know what was in the store at the place where the shopper lingered, you can’t attach meaning to the action. For in-store analytics, the basic context that gives meaning to the data is the store; specifically, what was there in the store at the place the visitor lingered. .... " 

Introduction to Robotic Process Automation (RPA)

Shades of expert systems, but with the ability to learn.   The integration of rule-based application of logic and analytic and deep learning methodologies.  Explicit and transparent expression of rules in play.  Including the ability to integrate a clear model of process.   Following.

Introduction to Robotic Process Automation (RPA)   35 page PDF

A Primer Developed and written by the Institute for Robotic Process Automation in association with Carnegie Mellon University

A Message from IRPA Founder Frank Casale: Hard Facts and Hype.

The Data and the Drama Behind Robotic Process Automation

Imagine a world in which the meaning of “work” has been redefined for millions of people. Where our service economy can actually focus on providing services, delivered by an engaged talent pool that is innovating on such service. In this new world, work would no longer be a “four-letter word” associated with functioning within repeatable systems and mundane transactional processes. 

Instead, this other world would have workers who rethink end-to-end processes on a more holistic level with the goal of simultaneously impacting several factors: quality, compliance, functionality, best practices, regulatory functions, customer satisfaction, human error, and the all-important bottom line – all while continuing to create “the next” in the form of remarkable products and services.  If you’ve come to this book, it’s probably because someone you know has told you that, in fact, this is the world we are already living in – that 2015 is to robotic process automation (RPA) what 1994 was to the Internet – an auspicious start, but we ain’t seen nothing yet! Thanks to RPA, we are well on our way to doing the business of the future. However, none of us can really predict just how revolutionary it will be. .... " 

Monday, July 24, 2017

Roomba Maps for Smart Homes

This reminds me of work we did with test Roombas.  We were interested in how they compared with human cleaning process and also usage of possible cleansing liquids, in lab settings.  But I can see how the resulting maps might be used for Smart Home navigation and understanding.  I recently read that the still unreleased Kuri robot automatically builds maps of its home space.  As expected, has this has privacy implications, mentioned at the link.

Your Roomba’s digital map of your house could be for sale     by Eric David in Siliconangle

 Robotic vacuums like the Roomba have made it easy for even the laziest people to keep their floors clean, and now Roomba maker iRobot Corp. has found another use for the little robots: building maps for your smart home devices.

Some of the smarter Roomba models build maps of your home to make their cleaning paths more efficient. Over time, they learn the locations of walls, doorways, lamps, furniture and so on, which eventually allows them to clean your home without bumping into things over and over. The idea behind this mapping system is to reduce the time it take for your Roomba to finish its job, but iRobot Chief Executive Colin Angle says the mapping data is useful for more than just speeding up your vacuum. .... " 

Swarms in Control

A favorite topic.  All systems are controlled by multiple entities.   Call that a swarm.   Stanford and their Platform Lab developing infrastructure for such systems.   Call it 'Big Control'.

Who Will Control the Swarm? 
A team of computer scientists and engineers are developing the infrastructure to centrally manage autonomous cars and drones ...
Stanford News   by Andrew Myers

Researchers at Stanford University believe future device swarms will operate via centralized management, using applications running in large data centers, similar to the way the cloud centralizes big data. In order to achieve this goal, the Stanford team has established the Platform Lab to develop infrastructure for these new "Big Control" applications. Although most current research into autonomous vehicles assumes a distributed model controlled in a peer-to-peer manner, with each machine doing its own calculations, a more concentrated model would have significant advantages, says Platform Lab faculty director John Ousterhout, who received the ACM Grace Murray Hopper Award for 1987, and the ACM Software System Award for 1997. Ousterhout notes this more concentrated model would provide a big-picture view of the world that enables better control of higher-level tasks such as system-wide situational perception, decision-making, and large-scale traffic planning. In addition, Ousterhout says the centralized applications can utilize powerful machine-learning algorithms that let the control systems learn and improve their behavior. The researchers are outlining a roadmap of the platform architecture. .... " 

Ignoring the Robots

Fascinating piece on the history of 'robots'.   Back to the 15th century.   Meant to do real jobs or just impress?  Far too expensive to just impress, but I will accept that is part of the reason.  ...... And we have long known that even the appearance of capability or intelligence can be a powerful thing.  But today there is much more attention paid to actually measuring value.  And their cost will not be ignored.

Please ignore the robots  in the Verge.
A lot of the robots we see are just for show — but the forces of automation are very real    by James Vincent    @jjvincent

Christopher Ahlberg Talks AI

Christopher Ahlberg, CEO at Recorded Future, and Staffan Truvé, Recorded Future’s chief technology officer, talks about AI in a Recorded Future podcast.

See more about Recorded Future:  Threat Intelligence Powered by Machine Learning 

Acquiring Digital Companies

Via Bain:

3 Ways M&A Is Different When You’re Acquiring a Digital Company  by Arnaud Leroi

Even for experienced deal makers, a first digital acquisition is bound to be an education. Companies acquire to accelerate their overall strategy and digital transformation, as Publicis Groupe did when it acquired Sapient for $3.7 billion in 2014 to help it make the leap from a traditional advertising company to a digital one. But when companies turn to mergers and acquisitions (M&A) to help them deal with digital disruption, they usually discover not only how different a beast digital M&A is compared with traditional M&A, but also that everything they thought they knew about M&A may actually not help. They’re also likely to be paying an even higher premium for the acquisition, betting on a fast—although uncertain—development.

Few executives appear prepared for the challenges of digital M&A. When we recently interviewed top M&A executives in Europe about their experience, fully three-quarters of them said that digital disruption has had a relatively large impact or even requires a complete overhaul of their M&A strategy. However, only 11% described themselves as being either “mature” or “advanced” on the learning curve. .... " 

Autonomous Cargo Ships Coming

Electrically Powered. But how they will address dangers like still common piracy is unclear.

The world’s first autonomous cargo ship will launch next year  By Duncan Riley in SiliconAngle. 

 The world’s first autonomous cargo ship is planned to be launched in 2018, as a company in Norway said Saturday that it has committed to build and deploy one.

Two companies, Yara International ASA and Kongsberg Gruppen, have joined forces to build the “Yara Birkeland,” a container vessel that will cart fertilizer on a 37-mile route from a production facility to a port in Norway. The vessel will have a capacity of 150 containers and will be electrically powered, meaning that it will also produce zero emissions.  .... "

Sunday, July 23, 2017

Machine Learning and Big Data

I was recently asked, whats the difference? ....  Just Architecture vs Math?

Machine learning with Big Data is, in many ways, different than "regular" machine learning. This informative image is helpful in identifying the steps in machine learning with Big Data, and how they fit together into a process of their own ...   

By Matthew Mayo, in KDnuggets.  (See the excellent chart, mentioned below, at the link) 

Big Data is no longer buzzword terminology or cutting edge, conceptually; rather, it just is. Big Data is not easily or precisely definable, but it is generally easy to identify when you see it.

While successful applications of machine learning cannot rely solely on cramming ever-increasing amounts of Big Data at algorithms and hoping for the best, the ability to leverage large amounts of data for machine learning tasks is a must-have skill for practitioners at this point.

While much of machine learning holds true regardless of data amounts, there are aspects which are the exclusive domain of Big Data modeling, or which apply moreso than they do to smaller data amounts. Data scientist Rubens Zimbres outlines a process for applying machine to Big Data in his original graphic below.  ..... " 

The Game of Checkers

Back on our early AI days, we looked closely at problem solving for games.  Notably the work of Arthur Samuel in machine learning.   This piece below on the seemingly simple game of checkers. Good about the people involved, but not enough about the nature of the AI.  See the last link for that, I apologize for the pay wall there.

How Checkers Was Solved
The story of a duel between two men, one who dies, and the nature of the quest to build artificial intelligence ..... " 

About Marion Tinsley, the world's greatest checkers player
by Alexis C. Madrigal in the Atlantic  .... 

Checkers was the first significant classic board game that had a non-human champion.  But it never defeated the long time human champion in a match.

See Checkers in Solved.  http://science.sciencemag.org/content/317/5844/1518  (Requires Login)

Amazon Chats

 Anytime by Amazon

Will customers use an Amazon app to chat?    by Matthew Stern in Retailwire. With discussion.

Smartphone messaging is perhaps one of the only areas of technology where we haven’t seen Amazon.com’s name show up front-and-center. But that may be changing.

Amazon is developing a messaging app called Anytime by Amazon, according to AFTVNews. Development appears to be in the very early stages, with Amazon now surveying customers about features they would look for in a messaging app.   .... "

Machine Learning and Hidden Decision Trees

Had posted about this once before.    We used decision trees effectively in the enterprise, and they can be useful due to their transparency.   In particular because it has a built in relationship to process.   Use it to work with decision makers to take next steps with a process improvement idea.  In DSC, revisiting hidden decision trees.     Also consider how you can use basic excel to do machine learning.

Saturday, July 22, 2017

Design for Users, not a Mythical Generation

The other day the question came up about how we should carefully design our user interface to match the generation of the current predicted user base.  Were they Gen X or Gen Y?   But have always felt this was a lame way to design. Consider the real needs first, and perhaps you won't need to fine tune the design at all.  Jimmie Lenz considers this:

Guest Opinion: Know Your Customer? Are You Sure? by Dr. Jimmie Lenz
July 21, 2017 The Financial Revolutionist  .... " 

Using Social Media to Find Top Customers

In Knowedge@Wharton.  

Influencing the Influencers: Using Social Media to Find Top Customers


Wharton's Gad Allon discusses his research on using social media data to identify a company's top customers.

Podcast/Text

Social media offers an almost endless stream of data for businesses to collect on their customers. But what good is data without a smart way to apply it? The latest research from Gad Allon, Wharton professor of operations, information and decisions, offers a lifeline for firms drowning in the deep waters of social networks. Allon and his team devised an analytics model that can help businesses identify high-value customers. The paper, “Managing Service Systems in the Presence of Social Networks,” was co-authored with Washington University professor Dennis J. Zhang. He talked with Knowledge@Wharton about the process. .... " 

Disney Tests Facial Expression Analysis for Movies

Back to using facial expression detection and coding.  Not new, Something we also experimented with, in lab settings, but here with new data science doing the analysis.  Some details in Phys.org.

Disney Uses Neural Net Technology to see if you Love its Movies like you Should  By Mark Coppock

Disney really wants to know if you like its movies. The entertainment company conducts copious research to answer that question, and in fact, it has its own division, Disney Research, that’s focused on figuring out ways to make you love its movies even more. Now, the company is applying neural net technology to determining if audiences are reacting as they should to Disney’s latest films.

The information comes via Phys.org,  https://phys.org/news/2017-07-neural-nets-audience-reactions-movies.html  which took a look at some recent work by Disney Research to apply deep learning concepts to figure out how audiences are reacting based on their facial expressions. So far, the methods are having better results than the organization’s usual techniques.

Researchers are using what they’re calling “factorized variational autoencoders,” or FVAEs, to use an audience member’s early facial expressions to predict how that person will react to a movie in its entirety. The data was gathered by using infrared cameras that monitored audience faces during 150 showings of a total of nine movies including Big Hero 6, The Jungle Book, and Star Wars: The Force Awakens. .... 

(Update)  Discussion in Fast CoDesign.   Disney Movies Could be Watching you. ... 

Origami Robots Moving Without Battery Power

Following uses of origami folding methods for simplifying robotics.   See much written before on this space regarding practical Origami, at the tag below.

Harvard's Folding Origami Robot can move without Requiring Battery Power     By Luke Dormehl   ....

Aspen Ideas on Future of Intelligence

The Future of Intelligence   Aspen Ideas Festival

Bill Gates portends doomsday is coming. Stephen Hawking says we should prepare for our robot overlords to take their thrones. But is the future as glaring as Hal’s red eye? Or is it more complicated than that? What does a future powered by algorithms and big intelligence mean for our lives? What are game-changing developments made possible by AI? What promises do these technologies hold? With responsible R&D, might we be looking at an application that can better the world? What are specific instances of AI’s promising power?

Speakers:
Gary Marcus, Tim O'Reilly, Michael Chui, Erik Schatzker ..... " 

Friday, July 21, 2017

Robot Physical Therapy

This came to mind in a recent physical therapy session, with my PT describing how muscles could be trained back into health   Why not hook up  a sensory means to watch the muscles,  detect changes in pattern,  and adjust the regimen towards a goal.  Here an example of work in this direction from EPFL. Though here about neurological changes and strokes, why not for any kind of physical therapy?

 In Newscientist: 
" ... Researchers at the Swiss Federal Institute of Technology in Lausanne (EPFL) and Lausanne University Hospital in Switzerland have developed a smart harness that uses artificial intelligence to help people regain their mobility following certain neurological injuries. The smart harness collects information on leg movement, stride, and muscle activity from body sensors, and then uses that information to provide support specifically designed for the way the patient walks, determining how much force to apply to produce a natural gait. Following a single, one-hour training session with the smart harness, patients with spinal cord injuries showed immediate improvement in their gait out of the harness compared to those who did not use the harness, according to the researchers. The system helps patients rebuild lost muscle mass and relearn posture and movement, while also retraining the brain to handle the balance between gravity and forward motion that walking requires. ... " 

Towards Lights-Out Fulfillment

'Lights-Out' used to be a favorite term when we worked in AI applications for the enterprise. Dramatically implying no humans, or lights needed to be involved.    Near total automation is still not here, but is approaching.  Consider the implications.  I note that Amazon now has 341K employees.  
Also it seems that most significant technological change mentioned here has Amazon involvement.   Now well over the tipping point for Influencing broad changes. Towards lights-out.  Keep watching their activity in this space.

Bezos move spurs $20B of growth in logistics robotics
Lights-out fulfillment will be the new norm within a decade   By Greg Nichols for Robotics.  In ZdNet.

This year is turning out to be a tipping point for how companies manage e-commerce fulfillment.

Worldwide sales of warehousing and logistics robots hit a respectable $1.9 billion in 2016. By 2021, according to a forecast by research firm Tractica, the market will hit a whopping $22.4 billion.

To better conceptualize the market explosion underway, consider that 40,000 units of warehousing and logistics robots were sold last year. A projected 620,000 units will be sold in 2021.

Why the change? In short: Amazon. If the last couple Prime Days proved anything, it's that a) Jeff Bezos is allowed to declare national holidays, and b) the investment Amazon made in warehouse logistics when it bought Kiva Systems for $775 million in 2012 presaged a complete transformation in global commerce.  ... " 

AI Exceeding Human Performance

Predictions are cheap,  machines have been faster than humans for some time, but when inserted in a set of tasks that that need to be done can be very different.   Thoughtful ideas here:

Intelligent Machines
Experts Predict When Artificial Intelligence Will Exceed Human Performance

Trucking will be computerized long before surgery, computer scientists say.  by Emerging Technology from the arXiv  May 31, 2017

Artificial intelligence is changing the world and doing it at breakneck speed. The promise is that intelligent machines will be able to do every task better and more cheaply than humans. Rightly or wrongly, one industry after another is falling under its spell, even though few have benefited significantly so far.

And that raises an interesting question: when will artificial intelligence exceed human performance? More specifically, when will a machine do your job better than you?

Today, we have an answer of sorts thanks to the work of Katja Grace at the Future of Humanity Institute at the University of Oxford and a few pals. To find out, these guys asked the experts. They surveyed the world’s leading researchers in artificial intelligence by asking them when they think intelligent machines will better humans in a wide range of tasks. And many of the answers are something of a surprise. .... " 

Thursday, July 20, 2017

Sears Integrates Alexa in Appliances, Sells on Amazon

Have noted earlier that GE is also integrating voice control in appliances.  Here an example of where a form of AI kicks up stock price, a rare event.    In Adage: 

Sears to Integrate Alexa Into Appliances and Start Selling on Amazon
Shares in Sears Holdings kicked off their biggest rally in almost two months after the company agreed to sell its Kenmore line on Amazon.com and integrate Amazon's virtual assistant Alexa into the appliances. That means the company's air conditioners and other devices will respond to voice commands.

Terms of the partnership weren't disclosed, and it's unclear how much of a boost Sears will get from the arrangement. But the company is badly in need of growth. Once the world's largest retail chain, Sears has racked up more than $10 billion in losses over the past six years.   ... " 

(Update)   Did Amazon just send Sears a life line with their Kenmore deal?   by George Anderson in Retailwire,  includes expert discussion.

If you can’t beat ‘em, join them. That appears to be the thinking behind an announcement yesterday by Sears Holdings that its Kenmore brand will now be sold on Amazon.com. ... 

The deal between the two companies calls for Amazon to own the inventory, but Sears will ship orders to customers from its warehouses. Innovel Solutions, a Sears business unit, will deliver and install the appliances.  .... " 

Amazon Spark Crowdsources Advertising?

Amazon takes another step forward.  I would imagine this will also emerge on the new Amazon Show screen as well.  A crowdsourcing of advertising to enthusiastic consumers?  Note their previous creation of an Influencer program.  Seems most every possible media channel is being investigated or even created by Amazon.

Amazon adds a shoppable “Spark” to Prime   By Deena M. Amato-McCoy in Chainstore Age.

Amazon Prime is getting more social as the online giant looks for yet more ways to entice shoppers to buy. 

On Wednesday, the online giant launched Amazon Spark, a new shoppable social media feed for its Prime members. Described as “a place to discover things from people who share your interests,” Spark will help Prime users discover — and shop for -- merchandise found across followers’ stories and ideas, according to Amazon’s website.

In a process similar to Instagram or Pinterest, Prime members use Spark to share photos of products they like. While users browse their Spark feed, they can tap on merchandise that inspires them and order that item instantly.   ..... " 

(more details at link)

Analytics Magazine

The latest analytics Magazine from Informs.    A practically oriented view of how analytics can be used in business today.   Industry News and Articles.

Limitations of Deep Learning

Just the introductory paragraphs to a post in the Keras Blog.   Worth following.   Useful statement of limitations,  on both narrow aspects of deep learning, and more generally about AI.   Technical, but readable as well.   I agree that the success of such models is quite unexpected, and to be considered with caution outside narrow applications.

The limitations of deep learning   By Francois Chollet

This post is adapted from Section 2 of Chapter 9 of my book, Deep Learning with Python (Manning Publications). It is part of a series of two posts on the current limitations of deep learning, and its future.

This post is targeted at people who already have significant experience with deep learning (e.g. people who have read chapters 1 through 8 of the book). We assume a lot of pre-existing knowledge.

Deep learning: the geometric view

The most surprising thing about deep learning is how simple it is. Ten years ago, no one expected that we would achieve such amazing results on machine perception problems by using simple parametric models trained with gradient descent. Now, it turns out that all you need is sufficiently large parametric models trained with gradient descent on sufficiently many examples. As Feynman once said about the universe, "It's not complicated, it's just a lot of it"   .... " 

(Update) See also, the Future of Deep Learning, from the same blog.  (technical)

Game Theory and Competitive Strategy

A favorite topic for corporate strategy considerations, but as we discovered, somewhat hard to apply in realistic circumstances.   Especially for competitive interactions. Fairly non-technical view.

In Game Theory, No Clear Path to Equilibrium   In Quanta Magazine.
In 1950, John Nash — the mathematician later featured in the book and film “A Beautiful Mind” — wrote a two-page paper that transformed the theory of economics. His crucial, yet utterly simple, idea was that any competitive game has a notion of equilibrium: a collection of strategies, one for each player, such that no player can win more by unilaterally switching to a different strategy.

Nash’s equilibrium concept, which earned him a Nobel Prize in economics in 1994, offers a unified framework for understanding strategic behavior not only in economics but also in psychology, evolutionary biology and a host of other fields. Its influence on economic theory “is comparable to that of the discovery of the DNA double helix in the biological sciences,” wrote Roger Myerson of the University of Chicago, another economics Nobelist.

When players are at equilibrium, no one has a reason to stray. But how do players get to equilibrium in the first place? In contrast with, say, a ball rolling downhill and coming to rest in a valley, there is no obvious force guiding game players toward a Nash equilibrium.  .... " 

How Alibaba Works with Retail Data

Via O'Reilly.

How Alibaba does retail data

What does it take to compete in a global market in which retail and the cloud are increasingly intertwined? Alibaba says it's domain-specific data science and machine learning for the masses. Here's how Alibaba's strategy differs from Amazon's—and how the company pulls it off.  

From ZDNet:  A view of Omnichannel.

Alibaba: Building a retail ecosystem on data science, machine learning, and cloud
Slide show By George Anadiotis for Big on Data     .... " 

Wednesday, July 19, 2017

IBM And Automation Anywhere Partner to Produce BPM Bots

I had been wondering for some time how IBM might be integrating business process management (BPM) and their cognitive methods to approach improving real business process.  Always seemed they had rule based systems, so why not?  Note the integration of software Bots.  Ultimately a very practical thing for improving operational management.   Powerful concept that we constructed piecemeal.  Don't like the term RPA, as it confuses the context.  Here from Forrester,  indications of some related steps: 

" ... IBM and Automation Anywhere (AA’s) today announced a collaboration (note-not a formal partnership yet) to integrate (AA’s) Robotic Process Automation (RPA) platform, used to create software bots to handle repetitive, task-based work, with IBM’s portfolio of digital process automation software, that includes IBM Business Process Manager and Operational Decision Manager. For AA, the partnership is a validation of its strong position in the RPA market, as shown in the Forrester Q1 RPA wave. For IBM, the partnership will enable clients to use AA’s RPA platform to create software bots that execute tasks within larger business processes managed by IBM’s software. Here’s our take.

IBM Can Add Smarts to AA’s RPA Platform

RPA works in a very dumb fashion today – mimicking human keystrokes and mouse movements – where all decisions must be explicitly programmed into the script. The result is that there are very few real decisions made beyond simple logic loops. Watson will be relevant down the road, but as a first step, RPA will benefit from IBM’s mature business rules engine (Decision Manager) or the embedded rules in the BPM platform. But as interesting, IBM’s content analytics, if part of the collaboration, can allow AA to keep pace with unstructured RPA intelligence from Workfusion and other RPA competitors moving quickly in that direction. RPA use cases that comb through unstructured content will be ahead of chatbot and AI-based digital workers. .... " 

Min Basadur's Simple Question

A former colleague of mine at Procter & Gamble,  Min Basadur's work is described.   Clever and very simple use of the question  "How Might we ... ?"  is mentioned.   Remember hearing of its use as a tool upon arrival at the company.  The very simplicity of the question makes it a great place to start.

Google and Facebook still use the 3-word question that saved a $225 billion company in the 1970s. 
... " 

Apple Journals its AI/Machine Learning Efforts

Apple has started blogging to draw attention to its AI work  : Sorry, it’s a ‘Journal’   by James Vincent@jjvincent in theVerge:

" ... The company’s new website, titled “Apple Machine Learning Journal,” is a bit grander than a blog. But it looks like it will have the same basic function: keeping readers up to date in a relatively accessible manner. “Here, you can read posts written by Apple engineers about their work using machine learning technologies,” says the opening post, before inviting feedback from researchers, students, and developers.

As the perennial question for bloggers goes, however: what’s the point? What are you trying to achieve? The answer is familiar: Apple wants more attention. .... " 

Looks to be very well worth following.

Claude Shannon

Great to see this article about Claude Shannon, who is little mentioned these days, but truly was a key element of understanding the value of knowledge in systems of information.  Had never heard the 'work habits' angle.

From Quora and Inc.
4 Brilliant Work Habits of History's Most Underrated Genius
Claude Shannon was the father of the information age, and we can all learn from him. .... " 

Google Glass Better for Business Applications

To put it simply,  Google Glass just looks more business-like than Augmented Reality Goggles.  It at least implies that you are still attached to solving problems in the real world.   Even better focused than you would be with AR.

Good piece in Engadget on the potential re-emergence in business.

Believe it or not, Glass may avoid being an all-time tech flop.
by Nathan Ingraham, @nateingraham  In Engadget.

" .... It was wild and impressive, but Google misjudged how that hype would translate into actual consumer usage. The look of someone wearing a camera on her face was too alienating, and Google never presented a complete vision of what Glass could do. That was part of the plan: The Explorer Edition that Google sold to early adopters was mostly meant for developers to use and figure out what apps made sense for it. But Glass never progressed beyond that experimental phase, and it was taken off the market in Jan. 2015, before a consumer edition even shipped.

Google never said Glass was dead, but it was clear the company's vision of a mass-market consumer product wasn't happening. So the company spent two years retooling and refocusing, and now Glass is back -- as an enterprise product meant to help workers get tricky jobs done. It's a rare example of parent company Alphabet significantly pulling back the ambition and scope of a product to serve a small market. In doing so, Google may give Microsoft's Hololens some unexpected competition in the augmented reality space. ... " 

CMO's Preparing for the Future

Former colleague of mine is interviewed. Business, now more than ever, is about prediction.

How CMOs Can Prepare For The Future: A Q&A With 'Predicting The Turn' Author Dave Knox   by Jennifer Rooney 

Dave Knox regards and engages with the marketing industry from an incredibly unique and insightful vantage point. As a veteran of Procter & Gamble, managing director of WPP Ventures, CMO of Rockfish, cofounder of The Brandery venture accelerator, startup advisor, blogger and author, he has a keen understanding of the power and challenges of big-name branding as well as the promise, innovation and opportunity of startup business.

His latest book, Predicting the Turn, provides CMOs with invaluable perspective on how to anticipate industry upheaval ahead and how best to prepare and lean in to those changes that can radically disrupt — even threaten — their businesses. Taking a page from startups, legacy companies can eschew the old habits that slow transformation and move forward for growth.  ... " 

On the Business of AI

Right now, for the second time in a long career, am involved directly in making AI part of a business.  So was struck by this view.

HBR:   On the business of AI. ...

".... What can AI do Today?

The term artificial intelligence was coined in 1955 by John McCarthy, a math professor at Dartmouth who organized the seminal conference on the topic the following year. Ever since, perhaps in part because of its evocative name, the field has given rise to more than its share of fantastic claims and promises. In 1957 the economist Herbert Simon predicted that computers would beat humans at chess within 10 years. (It took 40.) In 1967 the cognitive scientist Marvin Minsky said, “Within a generation the problem of creating ‘artificial intelligence’ will be substantially solved.” Simon and Minsky were both intellectual giants, but they erred badly. Thus it’s understandable that dramatic claims about future breakthroughs meet with a certain amount of skepticism.

Let’s start by exploring what AI is already doing and how quickly it is improving. The biggest advances have been in two broad areas: perception and cognition. In the former category some of the most practical advances have been made in relation to speech. Voice recognition is still far from perfect, but millions of people are now using it — think Siri, Alexa, and Google Assistant. The text you are now reading was originally dictated to a computer and transcribed with sufficient accuracy to make it faster than typing. A study by the Stanford computer scientist James Landay and colleagues found that speech recognition is now about three times as fast, on average, as typing on a cell phone. The error rate, once 8.5%, has dropped to 4.9%. What’s striking is that this substantial improvement has come not over the past 10 years but just since the summer of 2016. .... "

Multisensor Interfaces Reference

Of technical interest: 

Announcing: The Handbook of Multimodal-Multisensor Interfaces: Foundations, User Modeling, and Common Modality Combinations. Volume 1
Eeditors: Oviatt, Sharon; Schuller, Bjorn; Cohen, Philip R; Sonntag, Daniel; Potamianos, Gerasimos; Krueger, Antonio

Publisher: Morgan and Claypool/ACM Press.

https://lnkd.in/g93NPTD.

https://lnkd.in/gYPYKJU

This is a THREE volume series ..... "

Tuesday, July 18, 2017

Towards Innovation Cultures

Amazon, 3M, and Google All Have This Invisible Advantage
These five steps can help any organization gain an unseen edge over the competition.
By Soren Kaplan     Author, The Invisible Advantage  @sorenkaplan

Best Buy Wants to Install Smart Homes

Recall that Amazon just also announced a smart home consultation service, apparently aimed at getting their devices installed.  Probably a requirement for typical aftermarket installations of Smart Homes.  For the 'non digital native'.

Is Best Buy’s latest Geek Squad service a blueprint for niche IoT?  by Matthew Stern in Retailwire.  With discussion.

As Best Buy has worked to turn itself around, two of its biggest differentiators have been its Geek Squad service offerings and its focus on selling and supporting IoT devices. The chain’s latest move leverages both by focusing its Geek Squad resources on a very specific niche within the smart home/IoT area. Best Buy is offering smart home consultations and installations geared towards adults acting as caretakers for elderly parents.

The smart home service, called Assured Living, consists of a free assessment that allows caregivers to determine the right arrangement of in-home monitors, devices and alerts for effectively caring for the elderly people in question, the Star Tribune reported. Geek Squad employees then implement the technology. .... " 

Will start to cover these efforts on the term 'Smart Home Services' tag below.

Jumping Robotics

A small scale, jumping robot.  Quite impressive in details and related efforts.

Salto-1P Is the Most Amazing Jumping Robot We've Ever Seen,   By Evan Ackerman

Salto-1P uses a small motor and a system of linkages and gears to jump. Because it spends so little time in contact with the ground, the robot needs to do most of its control in the air. To do that, it uses a rotating inertial tail and two little thrusters to stabilize and reorient itself in between jumps..... "


Lucy from Equals 3

An example of Watson AI.

Superpower your marketing process with Lucy
Pull deep insights from piles of customer data. Make genius decisions. Compress months of work to one day.  ... Get big-brain assistance in research, segmentation, and planning.
Meet Lucy. Your ace time-saver — with A.I. skills.  (Uses Watson)

At your command, Lucy carries out vital tasks quickly with her superpowered information-gathering and analytic capabilities—freeing you for higher-order problem solving.

Lucy delivers brand Insights

Lucy is a voracious reader, digesting nearly 1,000 accredited news sources daily. That equates to roughly 250,000 articles a day! She then analyzes what she reads and delivers you up-to-date insights based on your brand mentions. .... "

Monday, July 17, 2017

Multiple Selves vs Better Bots

In Linkedin:

'Multiple Selves,' Not Better Bots, Is AI's 'Personal Productivity' Future

Published on April 3, 2017 Featured in: Big Ideas & Innovation, Productivity, Technology

by Michael Schrage,   visiting fellow at imperial college business school, innovation & entrepreneurship

MIT Sloan School's 'Initiative on the Digital Economy' recently published some of my new work on how the future of (inter)personal productivity inside the enterprise and out belongs to 'digital selves.' My essential argument: AI's true algorithmic impact will come more from 'Augmenting Introspection' than 'Artificial Intelligence.' We'll see 'Quantified Selves' emerging from the 'Quantified Self' as talented individuals and ambitious firms mash-up all manner of data and 'workplace analytics' in their collective quest to become more valuable and more productive. ... " 

EBay Looks at AI

We had some interactions with eBay in their early days.  Now Forbes does a quick look at how they are aiming to use AI methods to produce better focused personalized experiences.     Not a bad place to start, you already have a great deal of data, like customer scripts and resulting journies to start with.

More on the Google PAIR Effort: People+AI

 As important as how smart the methods will be,  will be the means by which we collaborate with them.   Thinking about this as a team, beyond just a person using a machine, will also be key.

Your Best Teammate Might Someday Be an Algorithm
A new program from Google seeks ways for AI systems to work more effectively with humans.
by Will Knight    In MIT Technology Review.  .... " 

(Update)  Wired writes about  'artificial stupidity' can arise.

Apple vs Google Maps

In Flowingdata:  

In Google and Apple mapping approaches.   I have looked at both of these for years now on the IOS smartphone.  A close comparative look at different mapping approaches.  Variation in data and design.   Choices in different user experience.  Shows the incredible investment and multiple redesigns that have been made.

Pharma and the Sharing of Data in Drug Discovery

Not quite what I call crowdsourcing, but rather a  means to make it easier to share and fund expensive research efforts.  Ultimately the very expensive genomic research efforts have to be funded too.   Good to see this direction being explored.

Big Pharma Buys into Crowdsourcing for Drug Discovery
By Menaka Wilhelm  in Wired

" .... Part of the problem is simply that drug design is hard. But many researchers point to the systems of paywalls and patents that lock up data, slowing the flow of information. So a nonprofit called the Structural Genomics Consortium is countering with a strategy of extreme openness. They’re partnering with nine pharmaceutical companies and labs at six universities, including Oxford, the University of Toronto, and UNC Chapel Hill. They’re pledging to share everything with each other—drug wish lists, results in open access journals, and experimental samples—hoping to speed up the long, expensive drug design process for tough diseases like Huntington’s. .... "

Machine Learning and Emergence

Complete article at the link in DSC:

The E-Dimension: Why Machine Learning Doesn’t Work Well for Some Problems?
by Shahab Sheikh-Bahaei, Ph.D.*
Principal Data Scientist,  Intertrust Technologies.

Introduction
Machine Learning (ML) is closely related to computational statistics which focuses on prediction-making through the use of computers. ML is a modern approach to an old problem:  predictive inference. It makes an inference from “feature” space to “outcome/target” space. In order to work properly, an ML algorithm has to discover and model hidden relationships between the feature space and the outcome space and create links between the two. Doing so requires overcoming barriers such as feature noise (randomness of features due to unexplained mechanisms).

In this article we argue that “Emergence” is also a barrier for predictive inference. Emergence is a phenomenon through which a completely new entity arises (emerges) from interactions among elementary entities such that the emerged entity exhibits properties the elementary entities do not exhibit. We present the idea that success of machine learning, and predictive inference in general, can be adversely affected by the phenomena of emergence. We argue that this phenomena might be partially responsible for unsuccessful use of current ML algorithms in some situations such as stock markets. .... " 

How Will People Interact with Machines

I see my former collaborators at IFTF have been working on this problem:

Forecasting How Humans Will Interact with Machines by 2030
By: Chris Preimesberger | July 12, 2017

The Dell-IFTF report forecasts that emerging technologies, supported by massive advancements in software, big data and processing power, will reshape lives and that society will enter a new phase in its relationship with machines. .... "

Sunday, July 16, 2017

Developing and Marketing Deep Learning

In Zdnet:

Automating automation: a framework for developing and marketing deep learning models 

Are you sold on the benefits of adding automation to your stack, but put off by the high entry barrier to this game? The NeoPulse Framework promises to ease the burden of developing Deep Learning models by introducing a number of interesting concepts.  .... "      By George Anadiotis

Assistant Mission Creep

When I read recently that a virtual assistant had called 911 in the US to report a crime in progress, my immediate reaction was:  No that will not work, for a number of reasons. Technical and regulatory. And that was confirmed soon after, it did not happen.  It was misreported.  But as AI's and their best known consumer examples, virtual assistants, continue to improve. our expectations will expand, leading to unexpected outcomes.  Will have to consider those circumstances and risk.  Nicely posed in Wired: 

“ ... Cyberservants will exhibit mission creep over time. They'll take on more and more functions. And they'll habituate us to become increasingly comfortable with always-on environments listening to our intimate spaces,” says Evan Selinger, a philosopher at the Rochester Institute of Technology, who focuses on how technology invades life.

SAVE's Reidenberg agrees, but thinks it’s inevitable that AIs will be given more power to intrude when necessary.“I don’t think we can avoid this. This is where it is going to go. It is really about us adapting to that,” he says.

Service Bot Examples

Systems that do sales bot interactions.  In cnbc:

" .... Here's how it works: When a company signs up with Conversica, they get to pick the name, gender and title of their new assistant. As leads come in, the AI assistant gets in touch with them through email or text message. If a lead is interested, the AI assistant routes the communication to a real-life member of the sales team to close the deal.

One advantage over humans is the AI isn't put off by unanswered emails — it doesn't mind being ignored or forget to follow up, so it can be programmed to be more persistent, emailing weeks after the initial contact. .... " 

Imagine a Future of Infinite Computing Power

Quite a way to go, but following how assistants can, learn, gather information, analyze and deliver.     I think its less about the speed of hardware and software, and more about how we model intelligent memory, data assets and interaction.

Amazon Imagines a Future of Infinite Computing Power
When David Limp thinks about the future of Alexa, the AI assistant he oversees at Amazon, he imagines a world not unlike Star Trek—a future in which you could be anywhere, asking anything, and an ambient computer would be there to fulfill your every need.

“Imagine a world in the not-so-distant future where you could have infinite computing power and infinite storage,” Limp said today at WIRED’s 2017 Business Conference in New York. “If you take off the constrains of servers and building up infrastructure, what could you do?”  ..... "

Saturday, July 15, 2017

Wal-Mart IOT Patent for Replenishment

Noted this before,  worth taking another look.  Note the implication of inclusion within the product.  Via CBInsights: 

Walmart’s IoT Patent Application Takes Aim At Amazon Dash

" ... The system is reminiscent of Amazon Dash — the connected buttons offered by Amazon that let users quickly re-order products. Launched over two years ago, Amazon Dash buttons now cover over 300 products. However, as we’ll describe below, Walmart’s system would require even less effort than Amazon’s. While Dash buttons still require users to press a physical button separate from the product (easy though that may be), Walmart aims to integrate IoT into the products themselves for automatic re-ordering with no user input at all. ... " 

Kasparov on Intelligent Machines

Kasparov talks from a unique perspective about machines and their intelligent use.

Don't fear intelligent machines. Work with them  Garry Kasparov TED Talk.

We must face our fears if we want to get the most out of technology -- and we must conquer those fears if we want to get the best out of humanity, says Garry Kasparov. One of the greatest chess players in history, Kasparov lost a memorable match to IBM supercomputer Deep Blue in 1997. Now he shares his vision for a future where intelligent machines help us turn our grandest dreams into reality.... " 

Whiskers as Sensors

Tactile sensors, like the idea of using new kinds of sensors.  Imagine the increased number of data points that could be gathered and fed to deep learning techniques.   Added to visual sensory input.

Want a Robot that can really feel?  Give it Whiskers.   By Matt Simon in Wired
" ... Whiskers are all the rage in nature, so why not give them to robots? Mechanical engineer Mitra Hartmann of Northwestern University is doing just that. In a new paper published in the journal Soft Robotics, Hartmann and her team detail how they’ve pulled one step closer to a rat-like machine that can feel an object and pinpoint it in 3-D space. Meaning robots of all kinds could soon get a powerful new sense. ... " 

How Will GE Adapt to Change?

Will be interesting to follow. They impressed me as a talented group, but also lumbering about as a big company without paying enough attention to good standards.  Will their Predix system take them in the right direction? In Knowledge@Wharton.