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

Thursday, April 19, 2018

Alexa Releases Skills Blueprint

OK, easy, and may kick up the Skill count.   But will it increase quality?   No code involved.   It might be better to make it easier to create plug-in capable micro skill-services that can be readily shared.  Not as easy,  as it depends on architecture to make that do-able.    Also note the implication that you might want to create your own private set of skills.  .Say for your home, family, company.   I like that idea.  Perhaps starting from a provided template.  Reaches closer to IFTTT involvement?

Amazon’s new ‘Alexa Blueprints’ let anyone create custom Alexa skills and responses   By Sarah Perez   @sarahintampa in TechCrunch.   

Amazon this morning is introducing “Alexa Blueprints,” a new way for any Alexa owner to create their own customized Alexa skills or responses, without needing to know how to code. The idea is to allow Alexa owners to create their own voice apps, like a trivia game or bedtime stories, or teach Alexa to respond to questions with answers they design – like “Who’s the best mom in the world?,” for example.

You could also create a skill that includes helpful information for the babysitter, which could be triggered by the command, “Alexa, open My Sitter,” Amazon suggests.

“Alexa Skill Blueprints is an entirely new way for you to teach Alexa personalized skills just for you and your family,” explained Steve Rabuchin, Vice President, Amazon Alexa, in a statement about the launch. “You don’t need experience building skills or coding to get started—my family created our own jokes skill in a matter of minutes, and it’s been a blast to interact with Alexa in a totally new and personal way.”  .... " 

Talk: AI and Business Process

Reminder, later this morning:

Talk by: Chitra Dorai,  IBM Fellow, Vice President and CTO  
“The Real World of AI: Business Processes Reimagined End to End”

19 Apr 2018: 10:30 AM, ET  Access Instructions below.

Talk slides and later the audio will be posted here when received.

Abstract: This talk will cover how enterprises are reimagining business processes with Al to bring about a frictionless  front office to back office transformation. Chitra will describe more than 30 cognitive solution use cases that we have  defined to change the way work is done today in Finance, HR and Procurement. I will discuss Al solutions deployment  experience, client case studies and business results that underscore how IBM is helping companies to embed Al in their  processes, and deliver end to end frictionless experiences and new outcomes 

Short bio: Dr. Chitra Dorai is an IBM Fellow and a Master Inventor, and is responsible for Cognitive Solutions and  Services in IBM's Global Business Services as its CTO. Dr. Dorai is a world renowned computer vision researcher and a  banking industry expert, and has received numerous awards and recognition both externally and at IBM for her  groundbreaking research and industry-transforming projects over the last twenty-seven years. 

Access:
Cognitive Systems Institute Group Speaker Series Thursdays 10:30am US Eastern Time
Join the meetings by pointing your web browser to:  https://zoom.us/j/7371462221 ; Callin: (415) 762-9988 or (646) 568-7788 Meeting id 7371462221 ; International Numbers: https://zoom.us/zoomconference.  .... 

Intel Backs off Smart Glasses

Perhaps a bit surprising given the enthusiasm they gave only a short while ago.    They had been working on object analysis as well.   So will we see new moves in this space?  I expect it from Asia.

Intel is giving up on its smart glasses    By Dieter Bohn in TheVerge

Intel has confirmed that it plans to shut down the New Devices Group (NDG), and cease development on the Vaunt smart glasses project we revealed earlier this year. The story was first reported this evening by The Information, which also notes that the closure will probably result in “some layoffs” from the team that was reportedly around 200 people.

Here’s Intel’s statement:

Intel is continuously working on new technologies and experiences. Not all of these develop into a product we choose to take to market. The Superlight [the codename for Vaunt] project is a great example where Intel developed truly differentiated, consumer augmented reality glasses. We are going to take a disciplined approach as we keep inventing and exploring new technologies, which will sometimes require tough choices when market dynamics don’t support further investment. ... " 

Microsoft Does Offline Translations

Looks impressive based on a demo I saw.   We are closer yet to a universal, mobile and free Babel Fish.  Was never expecting such a combination of features, the very basic idea seemed like fantasy not too long ago.  21 supported languages.  Offline packs for most popular languages.   Camera image text translation.   What other AI will be be carrying around soon?

Microsoft's AI-powered offline translation now runs on any phone
Translator should be more accurate when you're traveling abroad.

Jon Fingas, @jonfingas  in Engadget

Wikipedia Adds Preview Hover

Wikipedia adds a hover over page preview capability to make navigation easier.  Been testing, a nice idea.  Lets you make somewhat fewer clicks while you browse.

Navigating through Wikipedia articles on desktop just got a lot easier    By Olga Vasileva, Wikimedia Foundation ... More usage and testing stats ... 

Page previews, deployed today, is one of the largest changes to desktop Wikipedia made in recent years. .... " 

Wednesday, April 18, 2018

Your Data Has to be Good. Define Good.

Very good piece, covers lots of topics.  Worth reading.   Without enough quality data, you have nothing.  Key elements of process, like getting everyone involved early and often.  Biases mentioned, but the principle kinds of biases are not enumerated, and are often dependent on the business domain.  I like to count through likely biases and specifically test for some of the worst.

If Your Data Is Bad, Your Machine Learning Tools Are Useless    Thomas C. Redman in the HBR

Poor data quality is enemy number one to the widespread, profitable use of machine learning. While the caustic observation, “garbage-in, garbage-out” has plagued analytics and decision-making for generations, it carries a special warning for machine learning. The quality demands of machine learning are steep, and bad data can rear its ugly head twice — first in the historical data used to train the predictive model and second in the new data used by that model to make future decisions.

To properly train a predictive model, historical data must meet exceptionally broad and high quality standards. First, the data must be right: It must be correct, properly labeled, de-deduped, and so forth. But you must also have the right data — lots of unbiased data, over the entire range of inputs for which one aims to develop the predictive model. Most data quality work focuses on one criterion or the other, but for machine learning, you must work on both simultaneously.  .... " 

Will GDPR Drive us Towards Real-time Analytics?

Some good thoughts on the implications.   At its simplest will this mean we will need to use and release data more quickly?   Thus data will become real time?  At what cost of accuracy?

How GDPR Drives Real-Time Analytics  In DataFloq

New reforms under the General Data Protection Regulation (GDPR) started as an attempt to standardise data protection regulations in 2012. The European Union intends to make Europe “fit for the digital age.” It took four years to finalise the agreements and reach a roadmap on how the laws will be enforced.

The GDPR presents new opportunities as well as difficulties for businesses, digital companies, data collectors, and digital marketers. On the one hand, these regulations will make it more difficult for businesses and data mining firms to collect and analyse customer data for marketers, while on the other, they will present an opportunity for data collectors to innovate and enhance their techniques. This will lead to a better collection of more meaningful data, as customers will be directly involved. .... "

Machine Learning and Chaos

Quite remarkable, if the results can be applied to real world engineering problems.

Machine Learning’s ‘Amazing’ Ability to Predict Chaos by By Natalie Wolchover in Quanta Mag

In new computer experiments, artificial-intelligence algorithms can tell the future of chaotic systems.
alf a century ago, the pioneers of chaos theory discovered that the “butterfly effect” makes long-term prediction impossible. Even the smallest perturbation to a complex system (like the weather, the economy or just about anything else) can touch off a concatenation of events that leads to a dramatically divergent future. Unable to pin down the state of these systems precisely enough to predict how they’ll play out, we live under a veil of uncertainty.

But now the robots are here to help.

In a series of results reported in the journals Physical Review Letters and Chaos, scientists have used machine learning — the same computational technique behind recent successes in artificial intelligence — to predict the future evolution of chaotic systems out to stunningly distant horizons. The approach is being lauded by outside experts as groundbreaking and likely to find wide application.

“I find it really amazing how far into the future they predict” a system’s chaotic evolution, said Herbert Jaeger, a professor of computational science at Jacobs University in Bremen, Germany. ... " 

No indication in the paper abstract of how far such predictions can extend.   Or how their prediction value has been tested.     Technical:    https://arxiv.org/abs/1710.07313    A Challenge still it appears.

See also previous links to signal processing.


Developers Eye View of Smart Contracts

Its all about business rules.  So its about rule based process.  But why do it this way in particular rather than using rule bases and encrypted ledgers ?  Case is made here.  (Much more in full piece at the link)

Make your blockchain smart contracts smarter with business rules
By Stephane Mery and Daniel Selman  IBM

Blockchain already has a profound impact on multi-party business processes. Organizations that use blockchain count on trusted automatic transactions. It provides a framework of trust that you can use to securely automate processes that, until now, were often manual.

Traditionally, in business transactions, when two parties exchange value, they need to share a representation of the exchanged value and of the terms and conditions for the transaction. If they can't fully trust each other, each party maintains its own record of the exchange – a transaction ledger. They also keep their own copies of the rules and processes that govern the contract to exchange value.

Duplicating these records can lead to errors – and fraud. On top of this, the transaction processes are also duplicated, and performed manually and inefficiently. Disputes over discrepancies between the copies of the ledgers and contracts might need to be resolved in court, leading to significant costs, and ultimately delays to the transaction.

Blockchain uses cryptography along with distributed computing and storage technologies to solve the problems inherent in the traditional processes by providing trading parties with a means to share a trusted representation of their transactions and their assets.  ....  " 

Social Proof at the Wikipedia

Roger Dooley writes about the approach of the Wikipedia fundraising.  Many have seen their 'ads', as I have, when doing a Wikipedia search. The argument is  " ... Many people use our service, but few contribute ... "

From the Wikipedia: " ...  Social proof (also known as informational social influence) is a psychological and social phenomenon where people assume the actions of others in an attempt to reflect correct behavior in a given situation.   ... "

So you can argue in a way that says:  'Join all the many people who have contributed".  Or,  "Very few people contribute, so please do,  you can make a difference ... "

Wikipedia does the latter.  But suggests the former works better.

I have admit that I contributed based on this effort, and I rarely contribute based on online appeals.  So am I particularly anti-susceptible to this argument?

As a suggestion in the comments says, they certainly have the data and traffic to figure out which approach works better.   ... "

Holograms for Heads-up Displays

Have yet to see really practical and precise visualization via hologram.

Holograms Could Yield Next Wave of Heads-Up Displays
R&D Magazine    By Kenny Walter

A functional prototype heads-up display using holographic optical elements has been developed by researchers at the University of Arizona, a breakthrough that could increase the size of eye boxes used in heads-up displays for planes and cars. The new technology, if installed in a car, would enable a driver to see displayed information even if they moved around or was shorter or taller than average, says University of Arizona researcher Pierre-Alexandre Blanche. The researchers propose that laser light interactions could be used to fabricate holographic elements in light-sensitive materials that are smaller than traditional optical components and can be mass-produced. The new holographic elements redirect light from a small image into a piece of glass, where it is confined until it reaches another holographic optical element that extracts light. The researchers say the extracted hologram then presents a viewable image with a larger eye box size than the original image. ... "

Discovery Analytics Center

One former Alama Mater of mine has a discovery analytics center.

The Discovery Analytics Center brings together computer scientists, engineers, and statisticians to meet the research and workforce needs of today’s data-driven society.

About DAC

The Discovery Analytics Center at Virginia Tech is at the forefront of scientific innovation, leading Tech’s efforts in “big data” research and education on campus. We offer deeply technical undergraduate and graduate programs in analytics, and our faculty conduct leading-edge research in visual and text-based data analytics, machine learning, and computational statistics.

DAC brings together researchers from computer science, statistics, mathematics, and electrical and computer engineering to tackle knowledge discovery problems in important areas of national interest, including intelligence analysis, sustainability, and public health.  Our work emphasizes not just the algorithmic aspects of converting data to knowledge but also the importance of human-in-the-loop analytics to arrive at insights.

DAC was established in 2011 and has locations in Blacksburg, Falls Church, and Arlington.  Our team is comprised of 15 academic faculty members, 8 research and professional faculty, 3 administrative staff members, and 75 PhD students.  We are supported by the Department of Computer Science, the College of Engineering, Office of the Vice President for Research and Innovation, and the Institute for Critical Technology and Applied Science (ICTAS) at Virginia Tech. .... " 

Google and Innovation

Well put short piece. 

The 5 Essential Ingredients of a Truly Innovative Team, According to Google
You can't force creativity, but you can create an environment that encourages it. Google has figured out how. ... 
 
By Jessica Stillman in Inc

Tuesday, April 17, 2018

Alexa Assistants come to the Trading Floor

Another example where hands-free can be very useful, and you need information quickly.   Adding an information channel.

JPMorgan Brings Amazon’s Alexa to Wall Street Trading Floors
By Hugh Son and Katherine Chiglinsky  in Bloomberg Tech

Voice-activated assistant can now send reports from analysts

Other firms such as New York Life using it to help employees
“Alexa, ask JPMorgan what the price target for Apple is.”

It’s a request that JPMorgan Chase & Co. institutional clients can now get quickly answered through Amazon.com Inc.’s ubiquitous voice-activated assistant. The bank and the e-commerce giant have partnered to provide JPMorgan’s Wall Street users with another way to access its research. Alexa is able to send analysts’ reports and related queries, and the bank is testing other features, like providing prices on bonds or swaps, according to David Hudson, global head of markets execution for the New York-based bank.

Voice assistants are “clearly becoming something people are habituated to in their lives,” Hudson said. “It’s about taking information that’s somewhere in the bank, that someone has to generally go and look for, or which is time-consuming or requires authentication to get, and putting that to you in another channel.”  .... " 

Google Talks to Books

Google makes interesting use of their scanned books.  Could this also be used with other documents, say the knowledge of an enterprise?

Via Quartz:   (Many more good examples there)

Google’s astounding new search tool will answer any question by reading thousands of books

 " .... Imagine if you could gather thousands of writers in a circle to discuss one question. What would optimist Thomas L. Friedman say about intervening in Syria, for example?  Would chaos theorist Santo Banerjee concur?

Google now has a way to convene that kind of forum—in half a second. Speaking to TED curator Chris Anderson yesterday (April 13), legendary futurist Ray Kurzweil introduced “Talk to Books” a new way to find answers on the internet that should bring pleasure to researchers, bookworms and anyone seeking to expand their thinking on a range of topics.

Type a question into “Talk to Books,” and AI-powered tool will scan every sentence in 100,000 volumes in Google Books and generate a list of likely responses with the pertinent passage bolded. ... " 

 Google AI experiment has you talking to books  ...  in Engadget
The tech giant's other AI experiment is all about word association.

Mariella Moon, @mariella_moon in Engadget
5h ago in Internet  .... " 

Unintended Consequences of AI

Well put piece, we should be aware of the consequences of all forms of augmentation.  This one may be more hidden than most.

Tackling AI's Unintended Consequences     By Chris Brahm from Bain.

" ... How can companies address the loss of human expertise that accompanies the new freedom of artificial intelligence? Here are six risks for leaders to consider when making decisions on machine learning ... " 

"   .... As AI infiltrates more of our experiences and organizations, it’s important to recognize not only its many benefits but its unintended consequences as well. AI protects us from known and unknown threats, helps us connect to one another, and provides better answers faster and cheaper than humans do. And, of course, it’s great that AI frees us from routine tasks such as reading a map. But are we recognizing and addressing the loss of human expertise that accompanies that new freedom?

For business leaders and others investing in the technology, there are certain high-gain questions that can help them begin to grapple with leadership in the AI age—including how to manage the unique properties and risks of AI, bring clarity and focus to its deployment, and ultimately make better application of it (see Figure 1).  .... " 

Google Shopping Actions for Purchase

Was just asked to take a look at 'Shopping Actions' by a client.    Then saw this related piece on Retailwire.  Intriguing.   Connections to Google Home voice integration.   Indicates there might be a way to insert an assistant-style chat for product differentiation, engagment.  With further expert commentary.  Anything else out there?

Has Google found a formula for undercutting Amazon’s product search advantage?  by Tom Ryan in Retailwire

Google recently introduced Shopping Actions, a program that enables retailers to list products across Google Search, the Google Express shopping service and the Google Assistant app for smartphones and smart speakers like Google Home.

Mimicking the Amazon.com experience, shoppers browse and buy via a shareable list, universal shopping cart and instant checkout with saved payment credentials that work across Google.com and the Google Assistant.

“For example, shopper Kai can do a search on Google for moisturizing hand soap, see a sponsored listing for up & up brand soap from Target, and add it to a Google Express cart,” wrote Surojit Chatterjee, director of product management, Google Shopping, in a blog entry. “Later, in the kitchen, Kai can reorder foil through voice, add it to the same cart using Google Home, and purchase all items at once through a Google-hosted checkout flow.”  ...." 

Using Knowledge Studio

For the developer, a video showing how to use Knowledge Studio for a classic form of machine learning application.  Development-technical, but relatively straight forward example

Use Watson Knowledge Studio to build a custom machine learning model in the medical domain

About this webcast

One of the key benefits of building a machine learning annotator is the ability to train Watson in a complex domain such as medicine. Learn the methodology, standard practices, and guiderails on how to go about building an effective ML model. Steps include data understanding, type system building, pre-annotation, and deployment to WDS. After this session, you will have an acute understanding of what goes behind building an effective ML model. ... "

Alliance for Intelligence Driven Security

We connected with Recorded Future at an early stage.  Good to see their continued work.

Recorded Future Connect Xchange: Alliance for Intelligence-Driven Security

Integrations to Expand Impact of Traditional Security Operations, Incident Response, Security Orchestration and Automation, and Vulnerability Management Solutions With External Threat Intelligence

SAN FRANCISCO, April 16, 2018 /PRNewswire/ -- RSA CONFERENCE – Recorded Future, the leading threat intelligence provider, today announced Recorded Future Connect Xchange, a global technology partner community designed with one goal in mind: help security teams by enhancing their deployed tools through rich threat intelligence from outside their organization.

Connect Xchange partners will work together to increase accessibility to threat intelligence, allowing security professionals to proactively map applicable threats to their organization, and add dynamic, real-time intelligence to traditionally static security functions. Inaugural partners include: Brinqa, Cofense (formerly known as PhishMe), DFLabs, DomainTools, EclecticIQ, Farsight Security, IBM QRadar, IBM Resilient, LogRhythm, Palo Alto Networks, ProtectWise, ReversingLabs, Tenable®, ThreatConnect, and Versive.  ....

"Determining a security strategy without applying the lens of threat intelligence is akin to gauging the difficulty and danger of a hike based on what you can see from the trailhead. You know it's a mountain and the plan is to get to the top, but you don't know what predators are native to the region, how challenging the trail is, and so on. Understanding what's happening outside your organization, and applying that intelligence to your existing security solutions, is the only way to stand a fighting chance against today's attackers." 

— Dr. Christopher Ahlberg, CEO and Co-Founder of Recorded Future. .... "  

Monday, April 16, 2018

Alibaba Developing Driverless

Expect even more of this where there is less regulation.

Alibaba is developing its own driverless cars in MIT Technology Review

The Chinese tech giant comes to the self-driving game later than rivals Baidu and Tencent, but says it will build an entire ecosystem around autonomous cars.

The news: Alibaba confirmed today that it’s testing its own self-driving vehicle technology. The effort is led by Gang Wang, a scientist at the company’s AI lab and one of MIT Technology Review's 35 Innovators Under 35 in 2017.  .... "

(Updated) Talk: Linking Analytics to Models of Process for Frictionless Transformation

Talk this week a favorite topic of mine. About linking analytics directly to operational and strategic models of process.  Analytics now also means the use of cognitive and embedded logical capabilities.  Known broadly now as AI.   Have had several talks with IBM on this topic, including the linking of business process modeling to assistant technologies.    Contact me if you have any thoughts.  I will report on the talk.

19 Apr 2018: 10:30 AM, ET  Access Instructions below.

Talk by: Chitra Dorai,  IBM Fellow, Vice President and CTO  
“The Real World of AI: Business Processes Reimagined End to End”

Talk slides and later the audio will be posted here when received.

Abstract: This talk will cover how enterprises are reimagining business processes with Al to bring about a frictionless  front office to back office transformation. Chitra will describe more than 30 cognitive solution use cases that we have  defined to change the way work is done today in Finance, HR and Procurement. I will discuss Al solutions deployment  experience, client case studies and business results that underscore how IBM is helping companies to embed Al in their  processes, and deliver end to end frictionless experiences and new outcomes 

Short bio: Dr. Chitra Dorai is an IBM Fellow and a Master Inventor, and is responsible for Cognitive Solutions and  Services in IBM's Global Business Services as its CTO. Dr. Dorai is a world renowned computer vision researcher and a  banking industry expert, and has received numerous awards and recognition both externally and at IBM for her  groundbreaking research and industry-transforming projects over the last twenty-seven years. 

--------------------

Cognitive Systems Institute Group Speaker Series Thursdays 10:30am US Eastern Time
Join the meetings by pointing your web browser to:  https://zoom.us/j/7371462221 ; Callin: (415) 762-9988 or (646) 568-7788 Meeting id 7371462221 ; International Numbers: https://zoom.us/zoomconference.

Join the CSIG LinkedIn Group to get reminders about talks and discuss them. Use twitter: #CSIGNews & #OpenTechAI
Replays before Dec 2015:  Dial 877.471.6587 or 402.970.2667 and enter the call’s Replay ID when prompted for a program ID number.   The Replay ID is listed in the Recording column of each date.

Bird Calls from Audubon

A bit of a bird enthusiast myself, found this interesting.  It is quite simple. You recite the name of a bird, and you get back something like:  'There are 6 calls for this bird', and you choose one and it is played for you.   The audio is very good.  So the architecture is a simple look up, and in general the voice query gets it right.  But there is not much background to support the lookup,  and in a perhaps typical use scenario, you have heard a bird,  or seen an unknown bird, and want to look it up,  its hard to apply. Now if I could record a bird or take a picture of one and use that to search?

Amazon’s Alexa Is Ready to Help You Learn Bird Calls
The virtual assistant can now access more than 2,000 birds sounds from the Audubon library—as long as you say the magic words.  ... "

RFID File Tracking

Fascinating application of RFID tagging.   Reminds me of some of the challenges we had designing a much less challenging operation in retail.  Considerable detail, non technical:

Largest RFID File-Tracking System Goes Live in Qatar

The new system in place at Qatar Public Prosecution replaces an existing RFID solution to provide greater coverage with fewer readers, and to offer fast data capture, storage of data if a network goes down and real-time alerts in the event of unauthorized case file removals. .... 

By Claire Swedberg in RFID Journal

Common Sense - Project Alexandria

See my tag links below on our past looks at common sense based reasoning. Ultimately essential to get us to machine intelligence. 

Paul Allen Wants to Teach Machines Common Sense

Project Alexandria   By Cade Metz in The New York Times

Microsoft co-founder Paul Allen on Wednesday announced he would pour another $125 million into the non-profit Allen Institute for Artificial Intelligence to fund a project to teach computers common sense. He notes the additional funding should help to underwrite existing research as well as the common sense initiative, called Project Alexandria. Artificial intelligence (AI) "recognizes objects, but can't explain what it sees," says Allen Institute CEO Oren Etzioni. "It can't read a textbook and understand the questions in the back of the book." The Allen Institute wants to compile a database of fundamental knowledge that humans take for granted but which machines have always lacked, which will be fed into efforts such as Project Alexandria. "To make real progress in AI, we have to overcome the big challenges in the area of common sense,” Allen contends. .... 

Common sense is the everyday knowledge   that virtually every person has but no machine does.

More on AI2:

AI2 was founded in 2014 with the singular focus of conducting high-impact research and engineering in the field of artificial intelligence, all for the common good. AI2 is the creation of Paul Allen, Microsoft co-founder, and is led by Dr. Oren Etzioni, a world-renowned researcher and professor in the field of AI and computer science.

Situated on the shores of Lake Union, AI2 employs over 60 of the world’s best scientific talent in the field of AI, attracting individuals of varied interests and backgrounds from across the globe. AI2 prides itself on the diversity and collaboration of its team, and takes a results-oriented approach to complex challenges in AI.

AI2 has undertaken four main projects, Aristo, Semantic Scholar, Euclid and Plato to drive fundamental advances in science and medicine through AI. Meet our team, review our projects, and see our advisors and collaborators.

Based on these project efforts:

Alexandria,  Common Sense Knowledge and Reasoning
Aristo, Machine Reading and Reasoning
Semantic Scholar, AI-Based Search
Euclid, Natural Language Understanding
Plato, Computer Vision
AllenNLP,  Deep Semantic NLP Platform
Incubator,   Launching AI Powered Startups

Global Identity via the Blockchain

The idea of using blockchain for identity seems to be the clearest example of its use.  Had not seen Sovrin before, worth a look.

Via Sovrin, IBM supports user-centric global identity on blockchain

By James Kobelius in SiliconAngle

User-centric global identity is a dream that won’t die. What it refers to is a universal environment in which people can exchange self-issued digital credentials and rely on their legitimacy without the need for trusted third parties to vouch for and validate them.

This vision is a holy grail for libertarians and privacy advocates because it avoids third parties that might misuse, abuse or lose user identity information along the lines of Facebook Inc.’s recent Cambridge Analytica fiasco. In the early years of this millennium, Kim Cameron, who later joined Microsoft Corp., referred to the enabling architecture as an “identity metasystem” that is global, distributed and universally trusted. ... "

Knitting in 3D

Beyond the usual 3D approach.

Software Automatically Generates Knitting Instructions for 3D Shapes
Carnegie Mellon University
Byron Spice

Carnegie Mellon University (CMU) researchers have developed a system that can translate various three-dimensional (3D) shapes into stitch-by-stitch instructions executed by a computer-controlled knitting machine, and have used it to produce plush toys and garments. CMU professor James McCann envisions this milestone perhaps leading to on-demand machine knitting. His team's algorithm takes the knitting machines' limitations into account, generating instructions for patterns that work within these constraints and lower the risk of yarn breaks or jams. McCann notes additional work is needed to expand the system's capabilities beyond producing only smooth knitted cloth, which lacks the patterned stitching that can make knitted garments distinctive. "The software...needs a little push," McCann says, "and software can improve rapidly because we can iterate so much faster." The work will be presented in August at the ACM Conference on Computer Graphics and Interactive Techniques (SIGGRAPH 2018) in Vancouver, Canada. ... 

Sunday, April 15, 2018

A History of Advertising Holding Company WPP

Only did a little work with them on the edges of tech,  not very impressed, they seemed to be wrestling  with what they should do next.    Worth a historical read.

In Adage:

How WPP went from shopping cart maker to the world's Largest Advertising Holding Company

By Judann Pollack and Bradley Johnson. Published on April 15, 2018.

Getting the Most from your Data

Once more, Jason Brownlee makes excellent points.   It starts and ends with the data.   Its the biggest asset, and also the largest obstacle.  Frame it and test it.  Is there enough, does it support the methods you can use,  is it clean?  Here is his intro, much more at the link:


How to Get the Most From Your Machine Learning Data
by Jason Brownlee on April 16, 2018 in Machine Learning Process

The data that you use, and how you use it, will likely define the success of your predictive modeling problem.

Data and the framing of your problem may be the point of biggest leverage on your project.

Choosing the wrong data or the wrong framing for your problem may lead to a model with poor performance or, at worst, a model that cannot converge.

It is not possible to analytically calculate what data to use or how to use it, but it is possible to use a trial-and-error process to discover how to best use the data that you have.

In this post, you will discover to get the most from your data on your machine learning project.

After reading this post, you will know:

The importance of exploring alternate framings of your predictive modeling problem.
The need to develop a suite of “views” on your input data and to systematically test each.
The notion that feature selection, engineering, and preparation are ways of creating more views on your problem.

Let’s get started.... "

Panel: Retail and Machine learning

An interesting panel I am attending.  Like to see the idea of machine learning applied to retail, via Retailwire.

Machine Learning is Teaching Retailers How to Compete Again

Wednesday, April 18th, 1 pm Eastern / 10 am Pacific

Register and more information.  If you register you will get the panel recording.

A stellar panel, moderated by RetailWire's Al McClain.

Jared Brown    VP Data & Analytics, Tallan

Julie Bernard    Chief Marketing Officer, Verve

Phil Masiello      Founder and CEO, Hound Dog Digital Agency

Matt Kruczek     VP Web, Mobile and AI, Tallan

It's virtually impossible for retailers of any size to compete with the dominant online platforms based on pricing.

To maintain margins, SMBs are looking to differentiate in smarter ways — to carve out their own territory in areas rivals can't match. 

Recent advancements have made Machine Learning suites available to SMBs that can help them fortify their competitive strengths. We’ll cover ways retailers can start applying Machine Learning to:

Personalize recommendations based on learned intelligence 
Improve customer service through chatbot technology
Make customer "churn" predictions

Join us for a "humanizing" look at Machine Learning, including an engaging panel discussion exploring how this emerging tech can offer your business greater competitive opportunities.  .... " 

Robots not the Answer

But to some challenges they are inevitable.

Robots are not the answer to store challenges    by Nikki Baird in Retailwire with expert input.

... 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.

I love robots. In my spare time, I coach a high school robotics team. Robots are what will get us to Mars and on asteroids and inside volcanoes and to the depths of the ocean. They have an important place in industry too, accomplishing repetitive tasks with a high degree of precision and consistency.

But why do I keep hearing all these pronouncements about how robots are going to revolutionize the retail store – save it, even? Even before Amazon.com opened up its Go store and all the proclamations at NRF, lots of predictions heralded the rise of retail robots in stores. ... "

What Does an Agency Add?

Fairly good thoughts,  but not quite enough detail.   But the 'can ' and 'could'  must be rated by their return.   ...  but the potential return for digital could be large .... and what digital might  do,  might replace what agencies could do.  So what is left?  I do like mention of the processes involved, and agencies could be very good at that.

How an Agency Adds Value in an AI Era

By Rob Horler, Chief operating officer, Quantcast. Published on April 03, 2018.

" ... What will the agency of the future look like in this new AI era? How will agencies add value? As someone who spent almost 20 years in the agency business, I firmly believe in the value that agencies bring to their clients and I know there is an incredible opportunity ahead. However, this will require a disciplined focus on what agencies "can" do best versus what they "could" do. ... "

Machine Learning and Signal Processing

Quite interesting, linking classical signal processing and machine learning.   Signal processing was part of classical engineering training.   I assume this is still common.  Ultimately its all about signals that we are trying to detect, but here the signals contain random (Stochastic) noise and order we can extract.  This piece has coding examples, and is Technical, but straightforward in approach.   Worth at least filing away for problems of this type.

Machine Learning with Signal Processing Techniques By Ahmet Taspinar  Blog

Introduction

Stochastic Signal Analysis is a field of science concerned with the processing, modification and analysis of (stochastic) signals.

Anyone with a background in Physics or Engineering knows to some degree about signal analysis techniques, what these technique are and how they can be used to analyze, model and classify signals.

Data Scientists coming from a different fields, like Computer Science or Statistics, might not be aware of the analytical power these techniques bring with them.

In this blog post, we will have a look at how we can use Stochastic Signal Analysis techniques, in combination with traditional Machine Learning Classifiers for accurate classification and modelling of time-series and signals.

At the end of the blog-post you should be able understand the various signal-processing techniques which can be used to retrieve features from signals and be able to classify ECG signals (and even identify a person by their ECG signal), predict seizures from EEG signals, classify and identify targets in radar signals, identify patients with neuropathy or myopathyetc from EMG signals by using the FFT, etc etc.


In this blog-post we’ll discuss the following topics:

Basics of Signals
Transformations between time- and frequency-domain by means of FFT, PSD and autocorrelation.
Statistical parameter estimation and feature extraction
Example dataset: Classification of human activity
Extracting features from all signals in the training and test set
Classification with (traditional) Scikit-learn classifiers
Finals words  .... "

Saturday, April 14, 2018

Modeling the Visual Intelligence of Dogs

Quite a remarkable claim.   Can many visuals replace the ability to query a system about about what it knows.  Could this work with humans, entities like firms and their data?  Examining.

Via the University of Washington and Allen Institute for AI.  Who used neural networks to understand the behavior of dogs.   And suggest that animals provide data to train AI systems, including robotics.  What other visual agents?  And how accurate is the model?  Examining.

Who Let The Dogs Out? Modeling Dog Behavior From Visual   in arXIV
Kiana Ehsani, Hessam Bagherinezhad, Joseph Redmon, Roozbeh Mottaghi, Ali Farhadi

We introduce the task of directly modeling a visually intelligent agent. Computer vision typically focuses on solving various subtasks related to visual intelligence. We depart from this standard approach to computer vision; instead we directly model a visually intelligent agent. Our model takes visual information as input and directly predicts the actions of the agent. Toward this end we introduce DECADE, a large-scale dataset of ego-centric videos from a dog's perspective as well as her corresponding movements. Using this data we model how the dog acts and how the dog plans her movements. We show under a variety of metrics that given just visual input we can successfully model this intelligent agent in many situations. Moreover, the representation learned by our model encodes distinct information compared to representations trained on image classification, and our learned representation can generalize to other domains. In particular, we show strong results on the task of walkable surface estimation by using this dog modeling task as representation learning. ..."

Russia Has a Chat Assistant: Alice

 A global trend. More conversational? Yandex mentioned here previously. Russian language.

Alice In Putin’s Wonderland: How Russia’s AI Assistant Compares To Siri And Alexa
The assistant, launched by a company, with ties to the Kremlin, is exploding in growth and is not encumbered by any privacy or regulatory concerns.    By Daria Solovieva in Fastcompany

As artificial intelligence technology gets smarter and smarter, our concerns about user privacy keep growing. We’re asking why Amazon’s Alexa is laughing at users, whether Alexa can testify against you in court, and how Apple’s Siri can read your hidden messages out loud to other users. And Facebook’s privacy issues are getting so bad, the company has put its own plans for a smart speaker on hold.

But there is an AI assistant that shows no signs of slowing down and is not encumbered by any privacy or regulatory concerns. Launched in Moscow less than six months ago, Alice now has millions of daily users and “tens of millions daily interactions” according to Mikhail Bilenko, head of Machine Intelligence and Research at Yandex, Russia’s tech giant that runs the country’s dominant search engine and ride-hailing service.

Yandex launched Alice in October 2017, touting its first AI assistant as more conversational than its English-language competitors. Alice developers relied on the voice of a Russian actress who also provided the voiceover for Scarlett Johansson’s AI character in the Spike Jonze film Her. The AI assistant can be used by any Russian speaker anywhere on the planet.

“When it comes to non-informational chitchat, none of them (AI assistants) do it currently, there are a few pre-edited responses to certain questions, but that’s it,” Bilenko says in an interview with Fast Company. “Alice has a lot of directional, informational functions like the weather and currency (inquiries), but it also allows you just to chat about how your day was. The topic can be anything.” ... "  

Short, somewhat dated piece on Yandex.


(Update) Talk: Business Process and AI

Upcoming next week, a favorite topic of mine ... About linking analytics directly to operational and strategic models of process.  Analytics now also means the use of cognitive and embedded logical capabilities.  Known broadly now as AI.   Have had several talks with IBM on this topic, including the linking of business process modeling to assistant technologies.    Contact me if you have any thoughts.

19 Apr 2018: 10:30 AM, ET  Access Instructions below.
Talk by: Chitra Dorai,  IBM Fellow, Vice President and CTO  
“The Real World of AI: Business Processes Reimagined End to End”

Abstract: This talk will cover how enterprises are reimagining business processes with Al to bring about a frictionless  front office to back office transformation. Chitra will describe more than 30 cognitive solution use cases that we have  defined to change the way work is done today in Finance, HR and Procurement. I will discuss Al solutions deployment  experience, client case studies and business results that underscore how IBM is helping companies to embed Al in their  processes, and deliver end to end frictionless experiences and new outcomes 


Short bio: Dr. Chitra Dorai is an IBM Fellow and a Master Inventor, and is responsible for Cognitive Solutions and  Services in IBM's Global Business Services as its CTO. Dr. Dorai is a world renowned computer vision researcher and a  banking industry expert, and has received numerous awards and recognition both externally and at IBM for her  groundbreaking research and industry-transforming projects over the last twenty-seven years. 

--------------------

Cognitive Systems Institute Group Speaker Series Thursdays 10:30am US Eastern Time
Join the meetings by pointing your web browser to:  https://zoom.us/j/7371462221 ; Callin: (415) 762-9988 or (646) 568-7788 Meeting id 7371462221 ; International Numbers: https://zoom.us/zoomconference.

Join the CSIG LinkedIn Group to get reminders about talks and discuss them. Use twitter: #CSIGNews & #OpenTechAI
Replays before Dec 2015:  Dial 877.471.6587 or 402.970.2667 and enter the call’s Replay ID when prompted for a program ID number.   The Replay ID is listed in the Recording column of each date.

Monetizing Data to Produce Data as an Asset

A topic we followed and experimented with for years.   What has been added now is also the risk of holding data.   Consider risk analysis.   Good piece:

Monetizing Data: Follow the Yellow Brick Road   written by Mark Katz in FinancialTechnologyToday

Once Dorothy and her colleagues made the journey to OZ, they quickly found out that there was no there, there (every now and then, one must throw in some Gertrude Stein).  The Wizard simply told her what she really should have known all along.   That classic is about self-reliance, and the ability to differentiate, understanding the real value of what is already there–and the real range of possibilities.  Dorothy and her companions just had to figure out how to reframe their own perceived shortcomings and recast them as strengths to achieve real transformation.

Firms can undergo the same kind of journey, only to find out that there is indeed no “magic” to solving data monetization challenges.  While the tools have vastly improved, and the power of BI buttressed by AI and Machine Learning has helped greatly with incorporating challenges like unstructured and disparate data (internal and external), that Yellow Brick Road journey still requires cultural and operational steps including empowerment of associated teams.  There is not a lot of room for autocracy in achieving the best results.   Foundational elements work best, and collaboration is a must.   But ultimately, is it worth this kind of upheaval? ..... "

Getting the Most from AI

Attended this webinar last week from MITSloan, very well done. Here is a description and link to the recording.

On Demand: “Five Strategies for Getting the Most From AI”

Many business leaders are contemplating whether and/or how to introduce artificial intelligence into their organizations. The challenges of implementing AI are much discussed, but beyond implementation there is the far more important question: How do we generate competitive advantage from AI’s application?

In this webinar, Jacques Bughin, author of the MIT SMR article “Five Management Strategies for Getting the Most From AI,” discusses the work of the McKinsey Global Institute around this question of capitalizing on AI. Using industry examples and findings from the Institute’s research, he offers strategies for how to get the most out of AI’s potential.

In this webinar you will learn:

How to balance technical and talent investments
Why AI for growth is a better approach than AI for cost cutting
How to nurture the creation of AI ecosystems
How to align strategic goals with AI initiatives .... " 

AI Chips Advance Smarter Devices

Very good overview of the nature of chips that enable AI on the edge capabilities.  New capabilities for the IOT.  And who is doing what in delivering new hardware.  I recall mention back in the 90s it was said that faster hardware would be needed to make neural nets to drive solutions.  But at the time, there were few direct applications.   It was said then that software would drive solutions.   Now hardware is coming to the forefront.  Still expect software skills to be needed to communicate that power.

New AI systems on a chip will spark an explosion of even smarter devices  By James Kobielus in SiliconAngle

" .... That’s why mass-market mobile and IoT edge devices are increasingly coming equipped with systems-on-a-chip that are optimized for local AI processing. What distinguishes AI systems on a chip from traditional mobile processors is that they come with specialized neural-network processors, such as graphics processing units or GPUs, tensor processing units or TPUs, and field programming gate arrays or FPGAs. These AI-optimized chips offload neural-network processing from the device’s central processing unit chip, enabling more local autonomous AI processing and reducing the need to communicate with the cloud for AI processing..... "

Friday, April 13, 2018

Apple Homepod lagging?

Homepod will get many purchases simply from the wave of Apple zealots, and the audiophile speaker,  but is that enough?  Do the assistant aspects do well enough? Is that what people want?  How much is it connected to their Siri capabilities?  Expensive.  Good discussion:

Is Apple’s HomePod failing?
Two editors talk it out.
Engadget, @engadget

A report from Bloomberg earlier this week claimed that Apple's HomePod isn't doing so well, and that the company cut orders for new hardware from suppliers. This might not shock some of you: Apple missed the all-important holiday buying season and is competing with less expensive hardware from Google, Sonos and Amazon. But is the first smart speaker with Siri already a failure, or does the HomePod simply need time to find its place? ... "

New Amazon Skill Builders Guide

Amazon puts together a nice doc about building voice first experiences.  Linked to below.  Mostly rules of thumb to make things work.   And they have lots of data now on what works well, based on the architecture they have.    That may be the big win, continue to adapt approaches and architecture to lead to good assistant value for given contexts.  Note in particular the section mentioning the use of analytics to make enhancements.

10 Things Every Alexa Skill Should Do

Tips to Build High-Quality Alexa Skills

We get a lot of great questions from developers and skill builders about how to build high-quality Alexa skills. Questions like, “what makes a skill great,” “are there any best practices that I should be considering,” and “what can I do to make my skill better,” are usually top of mind.

Building Alexa skills is both an art and a science. There are technical aspects to building for voice, as well as creative concepts that go into designing natural voice experiences. Both are important things to consider as you build engaging skills.

With more than 30,000 skills in the Alexa Skills Store, we’ve learned a lot about what makes a skill great, and what you can do to create incredible voice experience for your customers.

Here are 10 things to keep in mind when you’re building voice-first experiences for Alexa: .... 

Your Brands Voice at Risk

Good cautions, but how different from other ad channels?

Don't Let Alexa or Siri Speak for Your Company: Protecting Your Brand's Voice on AI Platform

With the rising popularity of voice-driven AI assistants, it's important for businesses to offer a consistent brand experience. .... 

When companies first began establishing an internet presence a generation ago, brand-savvy businesses worked with designers and web developers to make sure their sites were customer-friendly and that content was consistent with their brand. They didn't just scan printed marketing collateral and post it online, leaving it to the search engine to present their material in the best light. But today, some brands are essentially taking the same risk by putting their brand's voice in the hands of third-party AI platforms.

That's a mistake.  .... "

Blockchain and the Supply Chain

Another example of the technology.   Similar it seems to what Wal-Mart has been experimenting with.  Still don't don't see why this is necessarily better than a well designed database, notably the speed aspect.

IBM Watson with blockchain boost adds visibility to supply chain disruptions  By Mark Albertson in SiliconAngle   ....A new, shared visibility ledger  Excerpt below with more detail at link.

" .... A key component of IBM’s focus involves its Supply Chain Business Network. This separate B2B connectivity offering has 6,000 clients, 400,000 trading partners and 8 million transaction documents per day, according to Suh. With that kind of engagement, it’s no surprise that IBM was interested in providing new technology solutions for its network of business clients.

The company took advantage of its conference in March to promote the benefits of a shared visibility ledger for its clients and partners in the Supply Chain Business Network. “We’re adding blockchain to that as a way to ensure transparency, as well as speed of operation,” Suh explained. “The shared ledger will allow you to see where in the process your transaction document is.”

IBM Watson enables “internet of things” devices to become active participants in transactions. The Watson platform processes and analyzes device-reported data or barcode-scan events. IoT-connected devices can talk with blockchain-based ledgers to validate smart contracts.

An example of this process in action can be found in the shipping industry. A.P. Moller–Maersk Group, one of the largest container ship operators in the world, recently formed a joint venture with IBM to digitize the global shipping ecosystem. Using the blockchain and Watson, the goal is seamless tracking of goods and shipments around the world, with updated paperwork in seconds instead of weeks.

“Small bits of optimization, meaning one percent improvement or resolving invoice and settlements, have such huge ripple effects downstream,” Suh said. “We’re excited because we’re now adding in not just AI capabilities, but also collaboration capabilities, which then allow groups of people to interact in-time and in-moment to address alternative decisions and routes.”

IBM’s application of cognitive processing to supply chain tracking is yet another example of the firm’s major bet on Watson. Suh’s organization is continually looking at new ways to integrate AI applications into business or consumer use cases. ... "

Voice Separation

Simple but powerful idea.  Have gone through voice training on several assistants.  This takes it further.  Note the idea of separating components of a conversation.

Google AI can pick out voices in a crowd
It could boost audio quality for video chats and hearing aids.

Jon Fingas, @jonfingas in Engadget
4h ago in Personal 

ComputingHumans are usually good at isolating a single voice in a crowd, but computers? Not so much -- just ask anyone trying to talk to a smart speaker at a house party. Google may have a surprisingly straightforward solution, however. Its researchers have developed a deep learning system that can pick out specific voices by looking at people's faces when they're speaking. The team trained its neural network model to recognize individual people speaking by themselves, and then created virtual "parties" (complete with background noise) to teach the AI how to isolate multiple voices into distinct audio tracks. .... "   Contains video and audio example.

More on Location Analysis

They may differ on the value, but its almost always done and reviewed closely.  Sure there are differences in its use among different kinds of retailers.  Consider at very least the capital investment involved.

Retailers Differ on Value of Location Analysis
by Brian Kilcourse 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.

RSR’s first benchmark on location analytics in retail showed there are significant differences in focus depending on what kind of products a retailer sells, and some of the differences are counter-intuitive.

Retailers across the different verticals, performance and size all agree that “more targeted marketing” is essential to their future well-being.

But why, for example, would almost twice as many fashion retailers as fast-moving-consumer-goods (FMCG) and general merchandisers (GM) rank “improved merchandising plan localization” as a high-priority? Fashion retailers introduce new collections every season and typically execute relatively few replenishments after the initial allocation is delivered to the stores. Grocers, drug stores and big box merchants, on the other hand, have many more SKUs in their standardized assortments and replenish very frequently. The payback from assortment localization based on customer-centered analytics are far greater.  ... "