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

Wednesday, September 19, 2018

Amazon may open up to 3,000 Cashierless stores by 2021

The data gathered must have produced some interesting results.  Now this would change the retail world:

Amazon May open up to 3,000 Cashierless Stores by 2021  in Adage

Sensors and Computer vision Technology

Amazon is considering a plan to open as many as 3,000 new AmazonGo cashierless stores in the next few years, according to people familiar with matter, an aggressive and costly expansion that would threaten convenience chains like 7-Eleven, quick-service sandwich shops like Subway and Panera Bread, and mom-and-pop pizzerias and taco trucks.

CEO Jeff Bezos sees eliminating meal-time logjams in busy cities as the best way for Amazon to reinvent the brick-and-mortar shopping experience, where most spending still occurs. But he's still experimenting with the best format: a convenience store that sells fresh prepared foods as well as a limited grocery selection similar to 7-Eleven franchises, or a place to simply pick up a quick bite to eat for people in a rush, similar to the U.K.-based chain Pret a Manger, one of the people said.   ... "

Talk: AI Fairness 360: Python Package

How to look at biases when your making AI driven decisions about people.

Invitation to the ISSIP Cognitive Systems Institute Group Webinar

 Slides and Recording: http://cognitive-science.info/community/weekly-update/

Date and Time: September 20, 2018 - 10:30am US Eastern
Talk Title: AI Fairness 360

Speaker: Kush Varshney, IBM
Talk Description:  
Machine learning models are increasingly used to inform high stakes decisions about people. Although machine learning, by its very nature, is always a form of statistical discrimination, the discrimination becomes objectionable when it places certain privileged groups at systematic advantage and certain unprivileged groups at systematic disadvantage. Biases in training data, due to either prejudice in labels or under-/over-sampling, yields models with unwanted bias.In this presentation, we introduce AI Fairness 360, a new Python package that includes a comprehensive set of metrics for datasets and models to test for biases, explanations for these metrics, and algorithms to mitigate bias in datasets and models. They have developed the package with extensibility in mind.  They encourage the contribution of your metrics, explainers, and debiasing algorithms. Please join the community to get started as a contributor.

Kush R. Varshney was born in Syracuse, NY in 1982. He received the B.S. degree (magna cum laude) in electrical and computer engineering with honors from Cornell University, Ithaca, NY, in 2004. He received the S.M. degree in 2006 and the Ph.D. degree in 2010, both in electrical engineering and computer science from the Massachusetts Institute of Technology (MIT), Cambridge.  While at MIT, he was a National Science Foundation Graduate Research Fellow.Dr. Varshney is a principal research staff member and manager with IBM Research AI at the Thomas J. Watson Research Center, Yorktown Heights, NY, where he leads the Learning and Decision Making group.  He is the founding co-director of the IBM Science for Social Good initiative.  He applies data science and predictive analytics to human capital management, healthcare, olfaction, computational creativity, public affairs, international development, and algorithmic fairness, which has led to recognitions such as the 2013 Gerstner Award for Client Excellence for contributions to the WellPoint team and the Extraordinary IBM Research Technical Accomplishment for contributions to workforce innovation and enterprise transformation. He conducts academic research on the theory and methods of statistical signal processing and machine learning. His work has been recognized through best paper awards at the Fusion 2009, SOLI 2013, KDD 2014, and SDM 2015 conferences. He is a senior member of the IEEE and a member of the Partnership on AI's Safety-Critical AI working group.

Date and Time : September 20 2018 - 10:30am US Eastern
Zoom meeting Link: https://zoom.us/j/7371462221
Zoom Callin: (415) 762-9988 or (646) 568-7788 Meeting id 7371462221
Zoom International Numbers: https://zoom.us/zoomconference
(Check the website in case the date or time changes: http://cognitive-science.info/community/weekly-update/ )

Please retweet  -

Join LinkedIn Group https://www.linkedin.com/groups/6729452

AI, Mixed Reality and Analytics Announced at Microsoft

Strong push at AI business applications and notably also their 'mixed reality' methods.   I still have not seen enough MR applications for the typical business. 

Announcing new AI and mixed reality business applications for Microsoft Dynamics  Alysa Taylor - Corporate Vice President, Business Applications & Industry

Today, I had the opportunity to speak to press and analysts in San Francisco about our vision for business applications at Microsoft. In addition, I had the privilege to make two very important announcements: the upcoming availability of new Dynamics 365 AI applications, and our very first mixed reality business applications: Dynamics 365 Remote Assist and Dynamics 365 Layout.

Our vision for business applications at Microsoft

We live in a connected world where companies are challenged every day to innovate so they can stay ahead of emerging trends and repivot business models to take advantage of new opportunities to meet growing customer demands.

To innovate, organizations need to reimagine their processes. They need solutions that are modern, enabling new experiences for how they can engage their customers while making their people more productive. They need unified systems that break data silos, so they have a holistic view of their business, customers and employees. They need pervasive intelligence threaded throughout the platform, giving them the ability to reason over data, to predict trends and drive proactive intelligent action. And with adaptable applications, they can be nimble, allowing them to take advantage of the next opportunity that comes their way.

Two years ago, when we introduced Dynamics 365 we started a journey to tear down the traditional silos of customer relationship management (CRM) and enterprise resource planning (ERP). We set out to reimagine business applications as modern, unified, intelligent and adaptable solutions that are integrated with Office 365 and natively built on Microsoft Azure.

With the release of our new AI and mixed reality applications we are taking another step forward on our journey to help empower every organization on the planet to achieve more through the accelerant of business applications. Specifically, today we are making the following announcements:  ... "
(more details here)

Mother of All Demos reprise

Those with interest in the history of modern computing know this event well. The 'Mother of all demos', done by Douglas Engelbart in 1968.   He drove home in a single demo what might be done with computers, without the user knowing any computer language.  Quite a new idea then.  Later we met and worked with Engelbart.  His later ideas dealt with collaborative, augmented problem solving.  Still a useful idea.
See also:  https://en.wikipedia.org/wiki/Douglas_Engelbart  

Watch this amazing demo of the early desktop computer
In 1968, computers got personal: How the “mother of all demos” changed the world.

By Margaret O'Hara in FastCompany

On a crisp California afternoon in early December 1968, a square-jawed, mild-mannered Stanford researcher named Douglas Engelbart took the stage at San Francisco’s Civic Auditorium and proceeded to blow everyone’s mind about what computers could do. Sitting down at a keyboard, this computer-age Clark Kent calmly showed a rapt audience of computer engineers how the devices they built could be utterly different kinds of machines–ones that were “alive for you all day,” as he put it, immediately responsive to your input, and which didn’t require users to know programming languages in order to operate.

Engelbart typed simple commands. He edited a grocery list. As he worked, he skipped the computer cursor across the screen using a strange wooden box that fit snugly under his palm. With small wheels underneath and a cord dangling from its rear, Engelbart dubbed it a “mouse.” .... "

Amazon Announces Gadget Alexa Tools

Been impressed by what Amazon is doing to support Devs.   Still nothing I would call breakthrough intelligence.    Waiting for that.

Amazon's new Alexa Gadgets Toolkit lets Alexa have some toy-friendly fun

Robots that can lip sync Alexa's speech? Disco balls that sparkle whenever you say the wake word? Amazon's new software makes it possible.   By    Ry Crist in CNET

Amazon has already made a point of making it as easy as possible for developers to build their own Alexa devices. Now, the online megaretailer wants to do the same thing for devices that work with Alexa -- and it wants to teach those gadgets some fun new tricks aimed at taking them to the next level.

To do so, Amazon is launching the Alexa Gadgets Toolkit, a new set of software tools designed for third-party gadgets that connect with Amazon's Echo devices via Bluetooth. Developers who put the toolkit to work will be able to take advantage of four new features. Here's how Amazon describes them: ... " 

In-Air Drone Charging

Involved in Drone tech examination and application.  Battery charging is a big issue.  Below article with video.   From IdeaConnection

GET In-Air Drone Charging Station  
The GET charging station for drones could offer unlimited flight time by letting the drones charge in mid-air.

The GET (Global Energy Transmission) company has developed an outdoor induction loop that can recharge several drones at one time without the need to land. The system, which resembles a wire frame, can transmit up to 12 kilowatts of power at an 80 percent efficiency to several drones simultaneously. A six-minute visit to the loop will provide 25 additional minutes of flight time, and the system is portable enough to be set up where needed—eliminating the need for human battery changers.   ... " 

Experience is the New Product

Experience Is the New Product; Here's How to Manage It
Organize around customer episodes, improving them through Agile teams.

By Gerard du Toit, Jens Engelhardt, Phil Sager and Karsten Fruechtl in Bain.

More companies now emphasize the entire experience surrounding a product or service. But how do they improve the experience, or fix a broken one?

For many, the key unit of management has become the customer "episode," and the core method is Agile. An episode consists of all the activities involved to successfully fulfill a customer's need.
While most Agile to date has focused on software development, improving an episode requires coordinating every factor that affects it, including product features, policies, processes, channels and technology.

Therefore, the team that owns an individual episode should comprise members from all the relevant functions. ....  "

IIOT Platforms

Interesting thoughts and visuals on the topic.

IIoT platforms: The technology stack as value driver in industrial equipment and machinery

Equipment and machinery companies considering a transformation to embrace the Industrial Internet of Things (IIoT) need to develop a clear perspective to drive impact at scale....

" ... Seeing the limits of hardware-driven growth, industrial equipment and machinery companies are looking to the Industrial Internet of Things (IIoT) to develop new customer-oriented, revenue-boosting business models. On the operations side, IIoT could increase production efficiency. Whether the focus is on revenue from new business models, savings from more efficient production, or both, digital-enabled advances in manufacturing require IIoT transformation. ...  "

In McKinsey

Tuesday, September 18, 2018

Cisco Talks Network Assurance with AI

Specific term was new to me.  But I do like the link to the business, the process, the goals.  Are the intents the same as in a simulation model of the business process?   Risks?  Reading more. 

Machine Learning for Analytics and Assurance
By Duval Yeager in Cisco BlogNetwork operation based in the intent of the business.  Goals?  Intriguing.

We are hearing amazing stories from our Cisco customers as they roll out intelligent analytics and assurance solutions in the form of Cisco DNA Center, Meraki insight, and Network Assurance Engine (NAE). The comments are on the accuracy and complexity of the analytical models that we have built, based on 30 years of Cisco networking leadership. You can read my blog post on how analytics works here. But, the back story to this is the approach of Machine Learning. When we add advanced machine learning algorithms to these products, the intelligence and system flexibility will be even more exciting. Let me explain…

Assurance in IP networking uses an analytics engine to verify that the network is operating based on the intents of the business. These intents are translated based on the network policies that IT configured when the system was set-up. The resulting model drives the decisions that an assurance solution makes to improve the network. This model is very good at network optimization, but every network is different, and network utilization is always changing as we change the way we use it  .... " 

Peter Norvig on the Breadth of AI Application

Links to recent O'Reilly talk.  Nice to hear Peter Norvig speaking on this, we read and used his early works on AI.  See: https://en.wikipedia.org/wiki/Peter_Norvig    AI is not just about neural pattern recognition, it can be much, much more.   Almost none of the companies I have surveyed realize this.  Have to believe his management at Google considerably upgrades their Breadth and Depth in AI.

Talk:  The breadth of AI applications: The ongoing expansion
Peter Norvig says one of the most exciting aspects of AI is the diversity of applications in fields far astray from the original breakthrough areas.

This is a keynote highlight from the Artificial Intelligence Conference in San Francisco 2018. Watch the full version of this keynote on O'Reilly's online learning platform.

You can also see other highlights from the event.  .... 

Peter Norvig
Peter Norvig is the Director of Research at Google Inc, where he has been since 2001. From 2002-2005 he was Director of Search Quality, which means he was the manager of record responsible for answering more queries than anyone else in the history of the world. He is a Fellow of the American Association for Artificial Intelligence and the Association for Computing Machinery and co-author of Artificial Intelligence: A Modern Approach, the leading textbook in the field (with 94% market share). Previously he was the head of the Computational Sciences Division at NASA Ames Research Center, making him NASA's senior computer scientist. He received the NASA Exceptional Achievement Award in 2001. He has served as an assistant professor at the University of Southern California and a research faculty member at the University of California at Berkeley Computer Science Department, from which he received a Ph.D. in 1986 and the distinguished alumni award in 2006. He has over fifty publications in Computer Science, concentrating on Artificial Intelligence, Natural Language Processing and Software Engineering, including the books Paradigms of AI Programming: Case Studies in Common Lisp, Verbmobil: A Translation System for Face-to-Face Dialog, and Intelligent Help Systems for UNIX. He is also the author of the Gettysburg Powerpoint Presentation and the world's longest palindromic sentence.   ... "

Uninformed Consent

Persistence of Surveillance online.   A non-technical survey of hidden capabilities and implications.

  Uninformed Consent  in the HBR

Companies want access to more and more of your personal data — from where you are to what’s in your DNA. Can they unlock its value without triggering a privacy backlash?

Three years ago the satirical website The Onion ran an article with the headline “Woman Stalked Across 8 Websites by Obsessed Shoe Advertisement.” Everywhere she went online, this fictional consumer saw the same ad. “The creepiest part,” she says in the story, “is that it even seems to know my shoe size.” The piece poked fun at an increasingly common — if clumsy — digital marketing technique. But today its gentle humor seems almost quaint. Technology has advanced far beyond the browser cookies and retargeting that allow ads to follow us around the internet. Smartphones now track our physical location and proximity to other people — and, as researchers recently discovered, can even do so when we turn off location services. We can disable the tracking on our web browsers, but our digital fingerprints can still be connected across devices, enabling our identities to be sleuthed out. Home assistants like Alexa listen to our conversations and, when activated, record what we’re saying. A growing range of everyday things — from Barbie dolls to medical devices — connect to the internet and transmit information about our movements, our behavior, our preferences, and even our health. A dominant web business model today is to amass as much data on individuals as possible and then use it or sell it — to target or persuade, reward or penalize. The internet has become a surveillance economy.    ... "

Efficiency for Machine Learning Automation

 And more elements of AI automation.  Here from MIT.   The details of the data construction for this is also described,  which is always enlightening.

Machine-learning system tackles speech and object recognition, all at once

Model learns to pick out objects within an image, using spoken descriptions.   By Rob Matheson | MIT News Office

MIT computer scientists have developed a system that learns to identify objects within an image, based on a spoken description of the image. Given an image and an audio caption, the model will highlight in real-time the relevant regions of the image being described.

Unlike current speech-recognition technologies, the model doesn’t require manual transcriptions and annotations of the examples it’s trained on. Instead, it learns words directly from recorded speech clips and objects in raw images, and associates them with one another.

The model can currently recognize only several hundred different words and object types. But the researchers hope that one day their combined speech-object recognition technique could save countless hours of manual labor and open new doors in speech and image recognition.

Speech-recognition systems such as Siri and Google Voice, for instance, require transcriptions of many thousands of hours of speech recordings. Using these data, the systems learn to map speech signals with specific words. Such an approach becomes especially problematic when, say, new terms enter our lexicon, and the systems must be retrained.

“We wanted to do speech recognition in a way that’s more natural, leveraging additional signals and information that humans have the benefit of using, but that machine learning algorithms don’t typically have access to. We got the idea of training a model in a manner similar to walking a child through the world and narrating what you’re seeing,” says David Harwath, a researcher in the Computer Science and Artificial Intelligence Laboratory (CSAIL) and the Spoken Language Systems Group. Harwath co-authored a paper describing the model that was presented at the recent European Conference on Computer Vision.

In the paper, the researchers demonstrate their model on an image of a young girl with blonde hair and blue eyes, wearing a blue dress, with a white lighthouse with a red roof in the background. The model learned to associate which pixels in the image corresponded with the words “girl,” “blonde hair,” “blue eyes,” “blue dress,” “white light house,” and “red roof.” When an audio caption was narrated, the model then highlighted each of those objects in the image as they were described. .... "

Related article.

DarwinAI to Automate AI Development

Another move towards the automating of AI, from U of Waterloo

DarwinAI launches from stealth to automate artificial intelligence development  By Maria Deutscher in SiliconAngle

 Artificial intelligence projects usually rely on existing code. Engineers sift through the AI models available from sources such as academic publications, pick the one deemed most suitable and then customize it to their requirements.

Canada’s DarwinAI Inc. wants to reduce the amount of work involved in the last two steps. The startup today launched from stealth with $3 million in funding and a specialized development platform that promises to automate AI projects.

DarwinAI’s software is based on research conducted by co-founder Alexander Wong at the University of Waterloo, where he serves as a professor with the systems design engineering department. The platform can customize an existing AI for the requirements of an application using its own built-in neural network.

According to DarwinAI, the software works by deriving a series of new models from the original based on user-defined parameters. Engineers can then pick the version with the particular combination of characteristics that best matches their target use case and, if necessary, make further enhancements. The startup said AI models created with its software are not only faster but also considerably more efficient.  ... " 

Monday, September 17, 2018

Facebook Advertising from Advertisemint

Just started to read Brian Meert's book:  'The Complete Guide to Facebook Advertising'   Nicely done.   Am no expert in Facebook advertising, but the book takes you through the process and motivations for using Facebook.    See also their company Avertisemint.    More to follow:

They Write:  " ... Facebook is one of the most popular social media sites to advertise one’s business, and for a good reason: it has more than two billion active users, a wealth of user data, and numerous ad formats. In the age of social media when nine in ten people check their phones an hour after waking up in the morning, Facebook advertising is the perfect complement to your advertising strategy. In The Complete Guide to Facebook Advertising, Brian Meert teaches you how to advertise on Facebook. He walks you through step-by-step guides filled with illustrations and easy-to-understand explanations. Additionally, he provides free resources and tips on how to create the perfect Facebook ad.  ... "   

Sisense to Understand and Visualize Data

Of interest. A challenge for the enterprise.  Will this solve it?  I like the idea of making the results broadly available in the enterprise.  Both for use of the data and understanding what you have available.  Now could it also intelligently suggest which data you need?   Point out the holes in your data, perhaps by industry context?

See also past posts on how Sisense can link with Amazon Echo ...

Sisense hauls in $80M investment as data analytics business matures  In TechCrunch By Ron Miller  @ron_miller

Sisense, a company that helps customers understand and visualize their data across multiple sources, announced an $80 million Series E investment today led by Insight Venture Partners. They also announced that Zack Urlocker, former COO at Duo Security and Zendesk, has joined the organization’s board of directors.

The company has attracted a prestigious list of past investors, who also participated in the round, including Battery Ventures, Bessemer Venture Partners, DFJ Venture Capital, Genesis Partners and Opus Capital. Today’s investment brings the total raised to close to $200 million.

CEO Amir Orad says investors like their mission of simplifying complex data with analytics and business intelligence and delivering it in whatever way makes sense. That could be on screens throughout the company, desktop or smartphone, or via Amazon Alexa. “We found a way to make accessing data extremely simple, mashing it together in a logical way and embedding it in every logical place,” he explained.   ..."

Data Imperative

Good O'Reilly Talk from recent conference:

The Data Imperative

Ben Sharma shares how the best organizations immunize themselves against the plague of static data and rigid process

By Ben Sharma September 13, 2018

BBC on Job Replacement

Jobs will be improved, switched in their nature, include many more cases of working with technology and both physical and cognitive robotics.  Be adaptive.  This will not mean you will have to learn advanced math or programming to be employed.

WEF: Robots 'will create more jobs than they displace'

Millions of jobs are likely to be displaced by automation but we have less to fear from robots than some might think, a report from the World Economic Forum has suggested.

The Swiss think tank predicts that robots will displace 75 million jobs globally by 2022 but create 133 million new ones - a "net positive".

It said advances in computing would free up workers for new tasks.

But others have warned there is no guarantee lost jobs will be replaced.

AI 'poses less risk to jobs than feared'
Bank warns on AI jobs threat
The WEF, which runs the famous Davos networking event, said that robots and algorithms would "vastly improve" the productivity of existing jobs and lead to many new ones in the coming years.  ... "

Customer Service State and Future

Good look at state and future of customer service, including advanced tech being used.   Have noted changes in how I use customer service.   In CustomerThink:

Customer Service: Where We Are and Where We Are Going
By Fara Haron in CustomerThink

Product and service-based businesses need a place to direct customers in need. If they don’t have one, it’s a fast road to an angry audience and a poor reputation. Customer service is a constant need across industries and has been for decades. However, not much light has been shed on the vast progress of this industry over the years. It started out with confused or upset customers picking up their landline for help, but today the average contact center supports an average of nine communication channels among email, web chat, social media and much more. To provide some insight on what’s happening behind the customer inquiry scenes, let’s take a look at the evolution of the contact center. How are customers currently serviced, how can we create the most efficient environment today, and how can customers expect to communicate 20 years from now? Perhaps more importantly, what is the data about customer expectations telling us? ... "

Advertising in the Age of Alexa

A former colleague writes. More at the link.

Advertising in the Age of Alexa or: How to Stop Worrying and Build Your Brand  By Lou Killeffer

Very glad to be returning as an Adjunct Professor in the School of Media and Journalism at the University of North Carolina at Chapel Hill.

When the "Advertising Campaigns" course I taught last year (Chobani & Chapel Hill) became unavailable, Carolina asked me to create a new course of my own, and last week the Curriculum Committee approved it: MEJO 490.5 "Advertising in the Age of Alexa or: How to Stop Worrying and Build Your Brand". 

My aim is an exploration of established advertising and brand theory and their evolving best practices in response to decades of continuous digital disruption. Through selected readings, engaging discussion, student research, and live interface with some of today’s most enlightened, real-world practitioners, we'll investigate  .... "

Modeling a Simple Organism

The nematode Caenor Elegans makes up about 80% of the animals on earth.  Yet it could be called the  most successful animal on earth.  Lives in most any environment.   We have completely sequenced its DNA.  We have been studying it for decades.  It has just 959 cells in its 'Brain',  302 neurons, each of its neurons are connected to 30 others.  So about 10K synapses.   (Humans have trillions).  So we should understand how a nematode's brain works?   No, not even close.   And we have been trying very hard. (Via Byron Reese in his book: The Fourth Age: Smart Robots, Conscious Computers and the Future of Humanity)  Good book on the direction, challenges and cautions of AI.

OpenWorm is an open source project dedicated to creating the first virtual organism in a computer.

Because modeling a simple nervous system is a first step toward fully understanding complex systems like the human brain.

By rejecting red tape and building a community of engineers, scientists, and other motivated volunteers from around the world. ... " 

Guide to Training Data

Nicely done and worth a look:

The Essential Guide to Training Data

A Guide for Machine Learning from Figure Eight
A machine learning algorithm isn't worth much without great training data to power it.

At Figure Eight, we've been providing that training data for a decade. We understand how to take raw data and annotate it so that it can be used to power the most innovative AI projects.

In The Essential Guide to Training Data, we share some of the lessons we've learned along the way. Download the guide and you'll learn about: 

Why simply using more data is often better than finding the latest cutting edge algorithm
Why just having a lot of big data isn’t the same as having labeled data
Where to find some great open datasets to bootstrap your model
Download the guide today.

Best regards,

The Figure Eight Team

Sunday, September 16, 2018

Intelligent Automation on Pace for Explosive Growth

Links of common processes to automation.   Good place to start before you decide to make all your systems 'intelligent'.    Note in particular coverage of RPA:  Robotic Process Automation.

Intelligent Automation on Pace for Explosive Growth, but Organizational Challenges Prevalent  by Kent Weare in InfoQ

In a recent KPMG study, the professional services organization published a report called Ready, Set, Fail?: Avoiding setbacks in the intelligent automation race which projects rapid growth of the intelligent automation (IA) domain. The report suggests that overall spend will reach $232 billion by 2025 compared to $12.4 billion which is spent today. But, this expected growth comes with many challenges, including tool maturity, skilled labor, organizational change management, governance and a lack of clarity involving return on investment.

Intelligent automation is an emerging set of new technology tools that mimic the actions a user would ordinarily perform to complete a task. Federico Berruti, a partner at McKinsey & Company, defines intelligent automation as: 

A suite of business-process improvements and next-generation tools that assists the knowledge worker by removing repetitive, replicable, and routine tasks. And it can radically improve customer journeys by simplifying interactions and speeding up processes.   ... "

P&G invents unique remedy for the common cold

From an un-expected direction by a  former employer.    It used to be that curing the common cold was an example of fabled innovation.   Does not seem to be that,  but a start?

P&G invents unique remedy for the common cold
BY Barrett J. Brunsman  – Staff reporter, Cincinnati Business Courier

Procter & Gamble Co. has been granted a patent for a liquid cold medication with the unique combination of a decongestant and the flavor of anise, an herb that some people find tastes similar to licorice or fennel.

The Cincinnati-based maker of consumer goods such as Vicks NyQuil cold and flu remedies (NYSE: PG) stated in a filing with the U.S. Patent & Trademark office that flavors are commonly added to liquid medications to mask an unpleasant taste.

Anise flavoring has been widely used in such medications, in particular nighttime multi-symptom relief cold and flu remedies, P&G stated. Many consumers not only enjoy that flavor but also trust and expect it. 

“Some consumers would like an anise flavored multi-symptom relief cold/flu medication that also contains phenylephrine hydrochloride, a decongestant,” P&G stated. “However, this formulation is not currently sold because it has surprisingly been found that anise flavoring (causes phenylephrine hydrochloride) to degrade, which makes the liquid medication less effective and significantly reduces the shelf life of the product.”  .... "

Saturday, September 15, 2018

Forrester Talks Digital Twins

Forrester pitches an expensive report below.   Worth a look.  Algorithms are interesting, but like with the term AI what they mean is usually more limited.  They work well in narrow domains for specific goals.  They do not mean they are intelligent nor are they even close to a 'Twin' of how we behave.  The term, like 'intelligence' can be used as a useful, broad description of a cognitive model of human capabilities.  It can replace some things that people do, but not all of them.  Again, beware the marketing.   Twin as a filter I like, so use the term filter.

The Algorithm Of You    By Fatemeh Khatibloo   Principal Analyst Forrester

Algorithm. It’s a buzzword you hear frequently these days. But does the average consumer understand the impact algorithms have on her life? Absolutely not. Consumers enjoy the illusion of unlimited choice in products, services, and content, but there’s almost always an algorithm behind the curtain, constantly refining and defining what’s presented to her.

Like it or not, we’re in the age of the algorithm.

Unfortunately, many businesses lost consideration for the human amidst the ceaseless data ingestion and optimization of these algorithms, and that’s led to negative outcomes for real people. While some governments are enacting regulation, we don’t think that solves the technology and ethics problem. Rather, we envision a new approach that puts the tools for data control in the hands of people and turns the tide of the impending “algopocalypse.”

In my latest report, “The Algorithm Of You: Meet Your Personal Digital Twin,” I introduce the concept of a personal digital twin (PDT), which I’ve defined as:

An algorithm owned by an individual, optimized for his or her personal objectives. It filters content that is counterproductive to the individual’s goals and identifies opportunities that support achieving them.

PDTs will eventually function as a digital proxy for each of us — they will learn what we value, what we prefer, and how we make decisions. We’ll be able to optimize them for hyperpersonalization, or serendipity. And as we do, the benefits will abound for the organizations we do business with. They’ll finally be able to:

Reach the right level of customer obsession.
Reduce marketing spending waste.
Optimize customer lifetime value.
Protect their brands from costly privacy and security missteps.
Build relationships with customers who actually want a relationship with them.
I’d love to hear your thoughts and invite you to reach out to me via inquiry or on Twitter!  ... "

AI Overhyped? The Term is.

The problem is the term AI itself.   The assumption that it is far more than it is.   This does not mean you should not think about how smarter capabilities could be inserted into codes to augment our capabilities.  For a while we were using the term 'Cognitive Systems' to indicate methods closer and even mimicking human perception and abilities.  Probably be better to ditch 'AI' and go with Cognitive.  Though even the latter requires too much explanation and can be over emphasized.   Our Cognitive Systems Institute, monitored here,  attempts to emphasize cognitive aspects. Beware over-marketing.

 Artificial intelligence is often overhyped—and here’s why that’s dangerous

AI has huge potential to transform our lives, but the term itself is being abused in very worrying ways, says Zachary Lipton, an assistant professor at Carnegie Mellon University.
by Martin Giles

To those with long memories, the hype surrounding artificial intelligence is becoming ever more reminiscent of the dot-com boom.

Billions of dollars are being invested into AI startups and AI projects at giant companies. The trouble, says Zachary Lipton, is that the opportunity is being overshadowed by opportunists making overblown claims about the technology’s capabilities.

During a talk at MIT Technology Review’s EmTech conference today, Lipton warned that the hype is blinding people to its limitations. “It’s getting harder and harder to distinguish what’s a real advance and what is snake oil,” he said.

AI technology known as deep learning has proved very powerful at performing tasks like image recognition and voice translation, and it’s now helping to power everything from self-driving cars to translation apps on smartphones,

But the technology still has significant limitations. Many deep-learning models only work well when fed vast amounts of data, and they often struggle to adapt to fast-changing real-world conditions.

In his presentation, Lipton also highlighted the tendency of AI boosters to claim human-like capabilities for the technology. The risk is that the AI bubble will lead people to place too much faith in algorithms governing things like autonomous vehicles and clinical diagnoses.

“Policymakers don’t read the scientific literature,” warned Lipton, “but they do read the clickbait that goes around.” The media business, he says, is complicit here because it’s not doing a good enough job of distinguishing between real advances in the field and PR fluff.

Lipton isn’t the only academic sounding the alarm: in a recent blog post, “Artificial Intelligence—The Revolution Hasn’t Happened Yet,” Michael Jordan, a professor at University of California, Berkeley, says that AI is all too often bandied about as “an intellectual wildcard,” and this makes it harder to think critically about the technology’s potential impact. ... " 

Alexa Wants to Simplify the Skill Experience

Some 50K skills, but many of them rather pointless. Now if only they could only work together to create something very useful. The unworldliness of all this will cause it collapse.   Not to say the basics of playing music, managing lights and cameras isn't useful and even fun.   But you can't just continue to chop the basic pieces up into smaller ones. 

Its not much different in Google Home.  There I found I could ask questions of the assistant in German,  but she insisted on answering with snippets in English.    I want things that can talk, converse, do useful things when they work together or with people.  Insightful and coherent, so I come out of the interaction a little smarter.   I would take a few hundred skills, if they were the right ones.  So when are we planning delivery?

Amazon wants Alexa to figure out how to fulfill users’ without help from Skills   By AJ Dellinger in Digitaltrends

Amazon’s wildly popular voice assistant Alexa has managed to make herself invaluable to many people because of the wide range of abilities available through Alexa Skills. Now, Amazon is reportedly getting ready to kill off the feature — well, kind of — in favor of a system that will simplify the experience for users.  ... " 

Friday, September 14, 2018

Watson Assistant Produces a Chatbot

IBM has announced a means of chatbot development, via their Watson Assistant infrastructure.  Recall I beta-tested the Dev under Watson Assistant.  The offer here, as I read it, is a one year free supported trial to built a chatbot using this approach.  The precise meaning of the 'chatbot' here is unclear, but I would assume it includes access to current Watson Assistant services.  Such as emotion detection  (anger) mentioned below.  Retail is mentioned but seems not restrictively.  Limitations of memory and access interactions. Looks like a good thing to try.

IBM Watson Assistant: the fast and easy solution for all your clients' cognitive needs.  Integrate an AI-powered chatbot at no cost for 12 months.

Create a cognitive retail chatbot

Learn how you can create an easily configurable, retail-ready Watson assistant-based chatbot that lets a user find items to purchase and then add and remove items from their cart.

See the code pattern

Create a cognitive banking chatbot

Use IBM Watson Node.js SDK to create a chatbot that includes conversation interaction, anger detection, natural language understanding, and answer discovery. ... 

See the code pattern

It’s estimated that chatbots for customer service will help businesses save $8 billion per year by 2022. ... "

Bringing in Information Theory

Nice idea.   Its really all about the knowledge we have now, and what we can add to that by learning.  And ways we can usefully measure that.  Now how can we leverage this idea?  This is ultimately a 'computer science' technical idea, via Information Theory, but the author makes it as comprehensible as is possible.  Intro below, and then off you go.

When Bayes, Ockham, and Shannon come together to define machine learning  by Tirthajyoti Sarkar in TowardsDataScience

Editorial Associate "Towards Data Science" | Sr. Principal Engineer | Ph.D. in EE (U. of Iilinois)| AI/ML certification, Stanford, MIT | Open-source contributor

A beautiful idea, which binds together concepts from statistics, information theory, and philosophy.


It is somewhat surprising that among all the high-flying buzzwords of machine learning, we don’t hear much about the one phrase which fuses some of the core concepts of statistical learning, information theory, and natural philosophy into a single three-word-combo.

Moreover, it is not just an obscure and pedantic phrase meant for machine learning (ML) Ph.Ds and theoreticians. It has a precise and easily accessible meaning for anyone interested to explore, and a practical pay-off for the practitioners of ML and data science.

I am talking about Minimum Description Length. And you may be thinking what the heck that is…

Let’s peal the layers off and see how useful it is…

L'Oreal Launches AR for Make Up

Again an area we experimented with for years, its seems to now be expanding.  Impressive and interesting details.  Makes a point that no App is required. This topic has been often mentioned here, tag below.  Lot of images at the

L’Oréal Paris Launches New AR Tool For Trying Out Make Up  in VRFocus
L'Oréal Paris and Modiface lets customers test make-up looks in AR with no additional apps required.

A number of companies and brands have begun using augmented reality to allow consumers to try out products and see how they might look in the home or about your person. There have even been a number of make up brands which have launched virtual try-on apps for testing out new looks, however most of them need the user to download a separate app. The new tool from L’Oréal Paris eliminates this requirement.

L’Oréal Paris have used facial recognition technology from ModiFace along with AR product simulation to allow users to discover and experiment with different make up and beauty products using only a web browser. ... " 

Decisions and Technology

Ultimately its about the how technological results translate into decisions.  So that makes this interesting.   Does tech enhance decisions, and how can you make it do that?  The influence of user interface?    Organizations, like a company or even an Army are ways that have been established with ways to influence decisions internally and externally.

Does technology really enhance our decision-making ability?     By ARL Public Affairs

ADELPHI, Md. -- An Army scientist recently won a best paper award at the Association for Computing Machinery's 26th Conference on User Modeling, Adaptation and Personalization for discovering that most people cannot distinguish between liking a user interface and making good choices.

Dr. James Schaffer, U.S. Army Research Laboratory scientist stationed at ARL West, and his collaborators at the University of California, Santa Barbara, Drs. John O'Donovan and Tobias Höllerer, received the best paper award at the conference held in July at Nanyang Technological University in Singapore.

So, does technology really enhance our decision-making ability?

The paper, "Separating User Experience from Choice Satisfaction," addresses this question and furthers the theory that underpins the evaluation of recommender systems, which are designed to help users make good choices.

Simply put, recommender systems are artificially intelligent algorithms that use big data to suggest additional products to consumers based off of things such as past purchases, demographic information or search history, for example. Think of the "people you may know" feature that exists on many of today's social media platforms.

In recommender systems, it has been assumed that users form very complex mental models of user interfaces.

This is reflected in current user experience measurements, which elicit subjective responses on a wide range of system features.  .... "

Risk:Challenging Probability

Kaiser Fung on the Book  ... Risk,    by John Adams  Which has some interesting thoughts about probabilities describing risk.  He begins:

I have been reading the excellent book by John Adams titled Risk (link). This is a geographer's treatment of a subject that is a staple of mathematics, particularly probability math. A mathematical treatment creates objects called probability distributions, which are then taken as complete representations of risk. Adams challenges that construct, bringing a social scientist's sensitivity to the table. In particular, he points out how the mathematics of risk is undermined by measurement issues (i.e. data issues) and statistical issues. He is not invalidating math, just pointing out large cracks that are often ignored.

I will provide a more comprehensive review of the book eventually. I'm very excited by Chapter 5, titled "Measuring Risk", and specifically the example of "traffic black spots". This example is very instructive for anyone who is interested in the practical implications and interpretation of risk measures.

The post got long so I have split it into two parts. The second part will be posted on Monday, and it concerns a delicious bit of analysis related to traffic black spots.   .... " 

Players in Supply Chain BlockChains

From an excellent presentation I participated in today by Dr. George Polak of Wright State on Blockchain in the Supply chain.

" ... Blockchain in Trucking Alliance (BiTA) comprises a 400+ firm consortium, including UPS, with the purpose of applying blockchain technology to logistics  ... 

Being developed by ShipChain (member of BiTA) to comprise a “fully integrated supply chain management system” based on blockchain. 

Would function as an open marketplace for logistics providers 

Shippers could track the capacity, cost and estimated delivery times for different routes for a given shipment before making a decision while carriers can dynamically adjust pricing based on supply and demand 

Would help to allocate resources effectively while avoiding markups by “rent-seeking brokers”
Source: “UPS bets on blockchain as the future of the trillion-dollar shipping industry,” by
Deep Patel, Bloomberg.com" . ... " 

Thursday, September 13, 2018

Neural Network for Snippets

A favorite topic, how do we make sense and value of written knowledge?

A Neural Network to Extract Knowledgeable Snippets and Documents 

Tech Xplore   By Ingrid Fadelli

Chinese Academy of Sciences researchers have created a convolutional neural network (CNN)-based model to extract knowledgeable snippets and annotate documents. The model can outperform current analytical tools while undergoing shorter training periods. The model is designed to comprehend the abstract concept of documents in different domains collaboratively and evaluate whether a document is knowledgeable, defined as one "containing multiple knowledgeable snippets, which describe concepts, properties of entities, or the relations among entities." The researchers say the network structure of their SSNN joint CNN-based model is "low-level Sharing, high-level Splitting," in which the low-level layers are shared for different domains while the high-level layers outside the network receive separate training to identify the differences of dissimilar domains. The team assessed SSNN's effectiveness on a dataset of real documents from three content domains on the WeChat messaging/social media/mobile payment platform. The model performed consistently better than other CNN models while saving time and memory usage due to shorter and more efficient training processes. In the future, the model could help build comprehensive knowledge databases and innovative services that answer user queries in real time. .... "

Microsoft Acquires Lobe:Deep learning without Code

Makes sense, this should now have progressed to custom no code solutions, with perhaps some minor tailoring to process needs.  Note the suggested breadth of communications services suggested.

Microsoft acquires AI startup Lobe to help people make deep learning models without code
By  Khari Johnson   @Khari Johnson

Microsoft today announced it has acquired Lobe, creator of a platform for building custom deep learning models using a visual interface that requires no code or technical understanding of AI. Lobe, a platform that can understand hand gestures, read handwriting, and hear music, will continue to develop as a standalone service, according to the company’s website.

People have only started to utilize the full potential of AI, Microsoft CTO Kevin Scott said today in a blog post announcing the acquisition.  ... "

Medical IoT

Lab makes data sharing easier so medical IoT devices can be smarter

Medical PnP laboratory researchers are looking to save lives through smarter, more interoperable healthcare technology using open standards, medical expertise and testing equipment that simulates health conditions to encourage easier integration of IoT devices and new sharing apps that can expand their capabilities.

By Jon Gold,  Senior Writer, Network World

AI Question Answering: Slides for Talk Today

Slides for talk today.  Recording will follow.

Sept. 13 10:30 AM ET Engineered AI Still Matters for Question Answering, By J. William Murdock, IBM

Zoom meeting Link: https://zoom.us/j/7371462221
Slides and recording will end up here:  http://cognitive-science.info/community/weekly-update/

Description: Many question-answering systems rely on a significant amount of engineering effort. They often require both knowledge bases and rules, which can be very expensive to create. Even when there is significant statistical machine learning involved in these systems, there is also an enormous amount of effort spent on identifying what features are useful for the machine learning and implementing capabilities (often using knowledge bases and rules) to assign values to those features. However, in recent years an alternative approach has been growing in popularity: single-strategy systems in which one statistical model is used to address the entire task. In this presentation, I will describe work in which we pursue both approaches and also integrate the two together. I describe results across two different data sets and show that purely statistical approaches are an excellent fit for some data, but that engineered knowledge and rules remain useful for more realistic and open-ended tasks. For additional details, see our paper at http://www.cogsys.org/papers/ACSvol6/article06.pdf   ..... 

IBM Talks Data Platforms for AI

An IBM blog post.   Agree with the issues, is this the solution?  Its about the data and the process being improved.   Choosing a Platform is a big deal.   By IBM:

Winning with AI
Since the year 2000, 52% of the companies that make up the Fortune 500 have disappeared. They have been acquired, succumbed to performance atrophy, or declared bankruptcy. In this hyper-competitive marketplace, winners and losers are being declared every day. And while artificial intelligence (AI) can be the valve to these pressures, for many, drafting a playbook for actually winning with AI remains daunting. 

Also, consider a recent IDC Cloud and AI Adoption Survey[1] in which more than 80% of respondents said they plan to move, or repatriate, data and workloads from public cloud environments to private clouds or on-premises locations over the next year, as the initial expectations of a single public cloud provider were not realized. These dynamics add to the confusion that every CEO, CIO, CTO, and CDO faces on a daily basis.

So, what precisely is dragging down projects and preventing companies from delivering measurable business value? I see three recurring patterns:

Companies have been accumulating data at an amazing pace for years, but are still challenged with how to store, manage, and control access. They need a new, modern approach;

The pressure to innovate is mounting. Companies create a chief data office or a data science center of excellence, but do not always have the right model for organizational success;

Small successes only scale when models are put into production and companies adapt their business processes, but unfortunately, this doesn’t occur very often. Scale requires platform thinking and technology.  Companies are at a critical juncture. They must be able to find and scale insights on demand if they want to climb the Ladder to AI.

Enter the Data Platform

This summer IBM launched an innovative approach and solution to this conundrum. Our new IBM Cloud Private for Data (ICP for Data) is a modern data platform designed to integrate data science, data engineering and application building into an environment that companies can use to uncover previously hidden insights from their data. Built on IBM Cloud Private, ICP for Data includes an enterprise meta-data catalog as the centerpiece along with services for data federation/virtualization, data warehousing, data integration, data science / machine learning and embedded dash-boarding.

Rob Thomas, General Manager, IBM Analytics.

It is designed to connect all data across an enterprise seamlessly, starting with enterprise data, and offers all its capabilities as data micro services. Consider it the highway system for the data revolution.  .... " 

New Apple Watch can do an ECG in 30 Seconds

Watched the detailed Apple presentation of this yesterday.  Amazing the whole thing can now be put into  a Watch and they claim the (one lead) ECG can be requested and be received in thirty seconds.    An AI analysis is said to do the analysis.  Can be recorded and sent to you doctor on a PDF file.    Some elements of 'diagnosis' are involved.   The watch is expense at $400 plus. See in Wired 'Apple Watch Could do more Harm than Good'   https://www.wired.com/story/ecg-apple-watch/     Some physicians at the presentation had similar objections.    ... 

Design Optimization, Simplification

Note these are optimization models with multiple objective goals, a very classic case.  We used this to look at various design contexts, like package design.    Visualization of results was used, even back then.

MIT Researchers Develop Tool to Simplify Product Design

Engineering.com  By Phillip Keane

Researchers at the Massachusetts Institute of Technology's (MIT) Computer Science and Artificial Intelligence Laboratory have developed a visualization tool for computer-aided design (CAD) that facilitates interactive, real-time exploration of design options that are best suited for sometimes competing performance goals. The InstantCAD tool combines multi-objective optimization methods with a CAD solution. MIT's Adriana Schulz says, "We're directly editing the performance space and providing real-time feedback on the designs that give you the best performance. A product may have 100 design parameters...but we really only care about how it behaves in the physical world." InstantCAD enables engineers to identify and convert the entire "Pareto front"—a set of designs optimized for all given performance objectives—into an interactive map. Clicking on the map displays optimized designs and variations in the immediate locale of that section of the front.

Short Summary of Smoothing Algorithms

In OpenDataScience.   A technical piece, but the descriptions are worthwhile to understand the why  as well.  Via O'Reilly.