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

Monday, October 15, 2018

Patents for Anomaly Detection

 There are many ways to do this kind of anomaly detection.  Done them for years.  Anomaly detection alone is not new or patentable, but perhaps as a larger process?

Anodot Gains Patents for Anomaly Detection By George Leopold in Datanami

Anodot, which focuses on using machine learning techniques to spot anomalies in time-series data, announced a pair of U.S. patent awards this week covering its autonomous analytics framework.

The analytics vendor said Thursday (Oct. 11) it has been granted two U.S. patents for algorithms that allow users to apply machine learning-base anomaly detection. The algorithms are designed specifically to quickly identify the source of anomalies in large data sets, then perform root-cause analysis. The approach is promoted as faster than traditional business intelligence tools or dashboards.

The first patent award covers a method for identifying and analyzing data anomalies by comparing them with previous incidents to “determine their sensitivity,” the company said. Anodot trains its machine learning algorithms based on human behavior rather than using statistical analysis tools.

“By leveraging machine learning and artificial intelligence capabilities, we’re able to tap into human perception and identify business incidents that other BI tools would never find,” claimed Ira Cohen, Anodot’s co-founder and chief data scientist.

The second patent award is for an algorithm used to identify “seasonal trends,” including daily and weekly patterns that could be used to improve detection of data anomalies. Anodot said the technology can be used to provide autonomous analytics alerts to business customers as incidents are detected..... "

TruStar and Enclave Defense

Newly brought to my attention by a piece in Recorded Future.  More pieces to follow.

Connective Defense is Here

TruSTAR’s intelligence platform uses Enclaves to streamline collaboration and provide enrichment from all of your trusted data sources, ensuring analysts have the right intel at the right time... all the time.   ....  "

Simulation for Autonomous Cars

Like the idea of simulation for difficult, rare or dangerous scenarios.   Of course we naturally do this  already when thinking:  "What would we do if?".   Automated simulation has existed for a long time, allowing us rigorously consider many possibilities at play.   The number of combinations at play (combinatorics) can be difficult.

Waymo's Cars Drive 10 Million Miles a Day in a Perilous Virtual World    In Technology Review  By Will Knight

Waymo, the self-driving subsidiary of Alphabet, says its autonomous vehicles have completed more than 10 million miles of driving since 2009. The company also said its software now drives the same distance inside a sprawling simulation of the real world every 24 hours—the equivalent of 25,000 cars driving non-stop—for a total of more than 6 billion virtual miles. The virtual test track enables researchers to examine the latest software updates on a variety of new scenarios, including situations that have not yet occurred in the real world. The system also makes it possible to test scenarios that would be too risky to do in reality, such as how the self-driving vehicle reacts when other vehicles are driving recklessly at high speeds. However, says the University of Michigan’s Ramanarayan Vasudevan, “The question is whether simulation-based testing truly contains all the difficult corner cases that make driving challenging.” ... ' 

Facebook has a Portal

More Detail on Facebook's Portal, will it compete with assistants?

Creepy or convenient? Facebook’s new Portal smart display follows you around the room

Get ready, folks, for a screened smart speaker from…Facebook?

Yep, you read that right. The social media giant today introduced two devices, which are both Amazon Alexa smart speakers with video displays. The Portal and Portal+ are hands-free devices with cameras and microphones for video calls to Facebook friends. As you might have guessed, a Facebook account is required.

If another person joins you on the call, everyone will always be in the frame.

“Both models are designed to make you feel closer to the important people in your life, and make video chats seem less like a call and more like you’re actually in the same room,” according to a Facebook press release.

The two models, offered in white or black, differ in size and functionality. The $199 Portal has a 10-inch display, while the $349 Portal+ boasts a whopping 15-inch pivoting display, which automatically pans and zooms on calls. That way if another person joins you on the call, everyone will always be in the frame. The camera and mics also automatically follow a caller moving around a room while they’re speaking ... "

Sunday, October 14, 2018

For Seeking Alternative Software

Sometimes you can't use the obvious software for a project.  Or don't know of alternatives.   Because you are building a cheap prototype under a budget, or you are a student, or you just want to see another example of how something can be done.

So you are looking for alternatives, their pros and cons, and specifications for their use.  I had a process simulation example I needed to address,  and needed to do something quick.  Was introduced to alternatives.me:

 ... Finding Better Alternatives    https://alternative.me
alternative.me is a platform for optimizing connections between a variety of software and product alternatives.  Currently we connected 8247 items. ... 

I then used this to look at alternatives to SimulationX, a process simulation platform.

 https://alternative.me/simulationx 

 ... SimulationX is a system simulation solution provided by ESI ITI. The company has been operating for over 25 years. The SimulationX software was introduced in 2015. Read more about SimulationX. .... 

My one concern was there were not enough professional reviews of the software examples, but still provided useful information for my purpose.

Google Assistant Re-designed

As more assistants get the ability to provide images easily.   Or show images on phones or displays when needed.

The Google Assistant gets more visual    By Frederic Lardinois   @fredericl in Techcrunch

Google  today is launching a major visual redesign of its Assistant experience on phones. While the original vision of the Assistant focused mostly on voice, half of all interactions with the Assistant actually include touch. So with this redesign, Google acknowledges that and brings more and larger visuals to the Assistant experience.

If you’ve used one of the recent crop of Assistant-enabled smart displays, then some of what’s new here may look familiar. You now get controls and sliders to manage your smart home devices, for example. Those include sliders to dim your lights and buttons to turn them on or off. There also are controls for managing the volume of your speakers.  .... "

Controlling Robotics with Brains

Though this creates many ethical implications, it is inevitable.  The title of this article is also misleading in emphasis, its more about controlling systems remotely than controlling soldiers' brains.

The Pentagon’s Push to Program Soldiers’ Brains  (Short excerpts from a long article) 

The military wants future super-soldiers to control robots with their thoughts. ... 

 .... The mission is to make human beings something other than what we are, with powers beyond the ones we’re born with.

 .... One aspiration: the ability, via computer, to transfer knowledge and thoughts from one person’s mind to another’s.

.... Will an enhanced human being—a human being possessing a neural interface with a computer—still be a human being? ...

Automated Farms

More progress in automated, robotic farming.  Lots of good images at the link.

Leafy greens are grown by machines at new, automated Silicon Valley farm in Digitaltrends

Did you hear the one about the Google software engineer who packed it all in to start a farm? No, it’s not the setup for a joke. Nor is it the premise for some quirky Sundance comedy, probably telling the story of a stressed-out programmer who rediscovers their happiness by moving to the country. It’s a real, honest-to-goodness farm, which just opened in San Carlos, around 20 miles outside San Francisco. Called Iron Ox, the farm aims to produce leafy greens — romaine, butterhead, and kale, alongside various herbs — at a rate of roughly 26,000 heads per year. Oh yes, and it’s staffed almost exclusively by robots.

“This is a fundamentally different way of approaching farming,” CEO and co-founder Brandon Alexander, 33, told Digital Trends. “Traditionally, the farming process means that you seed, you wait a few months, you come back, you harvest, and you distribute. That hasn’t changed a whole lot in hundreds, if not thousands, of years.” Until now, at least.

“This is a fundamentally different way of approaching farming”
Iron Ox’s indoor farm measures around 8,000-square-feet. That makes it paltry compared to the thousands of acres occupied by many traditional farms, but, through the use of some smart technology, it promises a production output that’s more in line with an outdoor farm five times its size. To achieve this, it has a few tricks up its sleeve. For starters, Iron Ox is a hydroponics farm, a method of growing plants without soil, using mineral nutrient solutions in a water solvent. Unlike a regular farm, hydroponic farms grow their produce in vertical and horizontal stacks; every element minutely controlled through the use of glowing LED lights and jets of water to affect the crops’ size, texture, and other characteristics.

In place of a farmer, Iron Ox employs a giant, 1,000 pound robot called Angus. It’s Angus’ job to move the heavy 800 pound, water-filled tubs of fresh produce without spilling them. A robot arm is used to tend the crops, making this the agricultural equivalent of Elon Musk’s automated Tesla factory in Fremont, CA.... " 

Saturday, October 13, 2018

The Keyboard, Now and Future

Long time interest in how humans and machines interact. And here is the most basic way.  Keyboard with a seemingly archaic design.   Why are we still using it?   What would it take to change? 

Why we can’t quit the QWERTY keyboard
We’ve been using it to type for 144 years. Here’s why it works, and what it would take for us to give it up.  by Rachel Metz  in Technology Review

 Tap is a one-handed gadget that fits over your fingers like rubbery brass knuckles and connects wirelessly to your smartphone. It’s supposed to free you from clunky physical keyboards and act as a go-anywhere typing interface. A promotional video shows smiling people wearing Tap and typing with one hand on a leg, on an arm, and even (perhaps jokingly) on some guy’s forehead.

But while Tap didn’t work well for me, it brings up an important question about the evolution of technology. We have so many ways to input data—by voice, touch screen, stylus, you name it—and yet we still rely heavily on something that looks a heck of a lot like the first commercially successful typewriters, released nearly 150 years ago. We can make powerful computers that fit in our pockets. So why haven’t we quit the keyboard? .... " 

KPMG on Intelligent Automation

Good thoughts here,  note mention of RPA, specific aspects of operations.  You should always learn much about your operational processes from any such effort.  Intelligent Automation.  Also the close involvement of decision makers at a level appropriate to the importance of the effort and likely changed involved.  Including C-Level.

 Ready, Set, Fail?: Avoiding setbacks in the intelligent automation race

New study reveals most organizations’ low readiness to deploy artificial intelligence technologies
Many traditional businesses with legacy approaches risk falling behind digital-first companies if they stay with the status quo. It takes a comprehensive transformation of business and operating models to compete in their own market at the level at which a Tesla or Amazon do in theirs.

Cliff Justice, KPMG Partner, Innovation & Enterprise Solutions, and leader of Cognitive Automation initiatives

Executives have high expectations for the impact of intelligent automation, but they're not yet ready to implement it from the top down and at scale. They'll struggle to get adequate ROI until they recognize two critical issues: 1) intelligent automation investment decisions need to be C-level strategy imperatives, 2) intelligent automation is about business and operating model transformation not simply technology deployment.

It's not clear whether most companies understand that intelligent automation is about changing business processes, and then restructuring the organization around those new processes now driven by technologies that didn't exist before. This means shifting the business and operating model from one of people supported by technology to one of technology supported by people. It's a digital-first operating model.                     

KPMG recently undertook a study to understand the reasons for and implications of deploying IA and what it takes to scale. KPMG professionals interviewed executives from numerous industries and geographies worldwide about their experiences with deployment and their perspectives on the future. Most emphasized that IA is poised to digitally transform their companies and industries and profoundly impact their employees' roles.                                                                           

At the same time, executives highlighted several challenges. In addition to grappling with the extraordinary pace of change, they are faced with understanding and choosing among hundreds of technology options, the need for effective data and analytics, prioritizing automation focus, and defining their future workforce. KPMG research considered three main areas of intelligent automation -- basic or robotic process automation (RPA), enhanced automation and cognitive automation.

These results underscore the need to not only act quickly but to plan deployments strategically with scale in mind. Most companies' executives acknowledged they are still experimenting only with RPA, applied to legacy applications and processes. With such a narrow focus and a bottom-up approach, they have not positioned themselves to transform their business and operating models so they can become and remain competitive with digital-first companies.     .... "

Blockchain and Proof of Work Efficiency

Ultimately efficient and secure architecture is key for block chain systems.

Bitcoin's Proof-of-Work Can Be Made More Efficient, IBM Research Claims  in CoinDesk

Scientists from IBM Research claim they've found a way to reshape blockchain architectures for energy-constrained internet of things devices. Read the full story ... 

Shared from Apple News ... "

Technology in Grocery

Like the loo at more mundane seeming aspects of grocery operations.

How 5 Top Grocers are Modernizing through Automation and Robotics
By Randy Hofbauer - 08/24/2018 in Progressive Grocer.

At Food Lion, Marty the Robot clean-sweeps the store a dozen times daily to identify slip-and-fall hazards on the floor, scan  shelves for out-of-stocks and ensure shelf pricing is aligned with the front end registers

As today’s top grocers seek to find ways to cut costs and allow their associates to focus more directly on attending to shoppers’ needs, they’re employing automation and robotics throughout the store, improving efficiency and accuracy in operations from food safety all the way to the last mile of delivery.

While many grocers have gone on record to share noteworthy ways that they’re working internally and externally to integrate new grocery technology in these areas — and even more are keeping their lips sealed — here are five noteworthy food retailers and the areas in which they’re employing the latest in robotics and automation solutions.... " ... 

Friday, October 12, 2018

Leadership Style and Getting People to Share

Is leadership stifling innovation and problem solving?   And a solution.

How Your Leadership Style Could Be Stifling Innovation and Problem Solving at Your Company  There's a reason fewer innovative ideas come up as your company grows.     Karin Hurt and David Dye write in Entrepreneur

Colleague Michael Schrage's HBR article is quoted:
A few years back, I helped a large, very compartmentalized and extremely silo-ed global organization launch an internal competition. Its goal was to promote greater sharing of ideas, information, best practice and innovative processes. Leadership recognized that business units and functions had effectively been allowed to ignore the rest of the enterprise. Significant opportunities and resources were left underexplored or untouched. They wanted to signal a cultural change but weren’t prepared to spend millions — or even hundreds of thousands — to achieve it.

The design was simple, clever and cheap: top management would recognize and reward people who demonstrated an ability to cross-functionally get real value from their colleagues and cohorts. We created two complementary yet competitive awards: “Thief of the Month” — a modest prize and high-profile internal acknowledgement for teams and small groups who “stole” an idea or innovation from another unit and successfully incorporated it into their own business; and “We Wuz Robbed” — a comparably modest prize and recognition for having one’s group’s best practice or process adopted by another internal group.

Dual prizes created a symmetrical “marketplace” where employees were simultaneously encouraged not just to look for interesting ideas to “steal” but to think about which of their own best practices deserved wider internal promotion. The competition thus incented both “supply” and “demand” of knowledge worth sharing. ... "

Fuzzy Algorithms for Selection

Interesting, especially with regard to fuzzy methods of classification ... Looking to see the details on this.   A way to produce a selection classification alternative.  Especially useful if the results can be simulated afterwards, maybe in some fuzzy way?

A self-portrait of NASA's Curiosity Mars rover. Software Finds the Best Way to Stick a Mars Landing   MIT News  By Jennifer Chu

Massachusetts Institute of Technology (MIT) researchers have developed software that could inform the selection of landing sites for scientific expeditions to Mars. The tool automatically generates maps of favorable sites based on available geology and terrain data, along with a list of scientific priorities and engineering limitations that users specify. MIT's Victor Pankratius and Guillaume Rongier created the software, and found it identified Mars landing sites that have been considered in the past, as well as highlighting other promising sites that were seldom suggested. The program depends on a probability-based, non-binary fuzzy logic scheme, and the team employs related algorithms to plot out initial favorability maps of possible sites gridded into individual cells. For each cell, the software calculates the probability that the site is favorable, producing a color-graded map and then applying a fast marching algorithm to chart paths a rover can take over a given terrain after landing. .... " 

Digiday: Bad Retail is Dead

Retail Briefing: ‘Bad retail is dead’

Retailers are facing a turning point: Sharpen up your data chops, or drop out.

As customers command more influence in retailers’ decision making today, these companies have to increasingly up the ante in order to succeed. Tools that were considered the business of tech startups not long ago, like AI, are now edging toward table stakes as retailers look to build out roadmaps that are shaped by customer demand. At NRF’s Shop.org conference in Las Vegas last week, conversation, both onstage and off, surrounded “smarter retail” strategies and tactics, including personalization, machine learning, predictive inventory buying, offline and online data analysis, responsive design processes and supply chains, localization and more.

“Retail isn’t dead. Bad retail is dead,” said CVS Pharmacy president Kevin Hourican onstage on Friday. Hourican laid out how every new development made at CVS is in some way in response to what customers have told the retailer they wanted. To turn analysis into action, CVS has mined data touch points, like how people flow through a store, what they’re buying in physical locations, and what they’re ordering through the mobile app. A few recent changes to CVS’s business include a new beauty department, next-day and same-day delivery and a push to make pharmacy processes more efficient with partner Aetna.  ... " 

Implications of Multiple Screens

This is not often talked, but it is important.   New kinds of channels to take advantage of this?

In Think With Google

In 2016, 85% of U.S. internet users surfed the web while watching TV.
 ‌
Source: "With a few tweaks, you can use the same ad on multiple platforms. Clinique shows how it's done" September 2017
 ‌
Why it matters
 ‌
A multiscreen world poses a challenge for brands. After all, creating separate campaigns for each platform can stretch even the healthiest of marketing budgets. What if instead of reinventing the wheel for each digital campaign, you could transform your print ads into six-second bumpers by adding motion graphics, saving creative resources, time, and money? Clinique showed how it's possible to do more with less.  ...  " 

Thursday, October 11, 2018

Colgate Rolls Virtual Reality in Plant and Office

Note the mention of manufacturing/maintenance use as well as uses for back office.  Impressive rollout for multiple areas.

RealWare Rolls Out AR Wearables At Colgate-Palmolive

The RealWear HMT-1 AR headset is being deployed at Colgate-Palmolive facilities.  By Rebecca Hills-Duty in VRFocus

Augmented RealityColgate-PalmoliveIndustryRealWear
A number of industries and companies are turning to augmented reality (AR) headset technology to improve workflows and provide workers with information while leaving the hands free. Colgate-Palmolive has decided to work with wearables company RealWear to introduce a voice-operated HMT-1 AR headset to its workforce.

Using the new technology, Colgate-Palmolive employees will be able to troubleshoot machinery while keeping hands free to work, using voice to talk to company experts, equipment suppliers and manufacturing teams.

RealWear have created the HMT-1 headset to be a hands-free, hard-wearing head-mounted device which is suitable for use in a variety of industrial locations and industries, including oil and gas, utilities, automotive and manufacturing.

“Colgate-Palmolive tested RealWear HMT-1 through successful pilots in 8 locations and is now standardizing globally on this wearable device for our manufacturing operations,” said Warren Pruitt, VP Global Engineering at Colgate-Palmolive. “Looking ahead, we see an opportunity to use this tool beyond the plant floor for improved performance and new efficiencies.”

“Colgate-Palmolive’s global deployment signifies that industrial wearable computing has crossed the chasm into mainstream manufacturing companies,” said Andy Lowery, RealWear CEO and cofounder. “This has been a textbook example of how to move with energy and purpose, from evaluation through pilot to global deployment.”

The company is also planning to use the devices to capture and retrieve information on documents and show informational or training videos. The RealWare platform can offer service relating to remote mentoring, document navigation, internt of things (IoT) visualisation and digital workflow solutions.  .... "

Extended Google Smart Home Ecosystem

Just today I extended out smart home ecosystem, which uses both Google and Amazon, to another room.  My interest is always in ease of installation, use,  extension and maintenance.   And the resulting augmentation of the home and home office user.   Things are always getting better, but still not completely transparent for the average home do it yourself user.  Often this depends on other infrastructure.  Like wifi, electrical system, manufacturers of basic installed devices.

Google's smart home ecosystem has arrived
It's not as robust as Amazon's, but it's a strong competitor.
 By Devindra Hardawar, @devindra    In  Engadget

The building blocks of the "smart home" -- connected lights, plugs and other gadgets -- have been around for years. But they were strung together by a variety of confusing protocols. It wasn't until Amazon's Echo and its accompanying Alexa virtual assistant came along that the smart home started to make sense. We didn't need just smart devices; we also needed voice controls and seamless interoperability between devices. Basically, we needed a proper smart home ecosystem.  ... "

Who you Need in Your Network

I like the idea, but it does depend very much who they are and what they are willing to contribute.  Even the case with actual boards of directors.   Also, keep your network well connected, not usually the case even these days

The 5 people you must have in your network  In  FastCompany.com 

The term “mentor” dates back to Homer’s Odyssey. But the world of work has changed, and you no longer need a mentor: You need your own personal board of directors.

The role and impact of mentors may vary, but the generally accepted business practice is that people need a mentor. It’s time to challenge that wisdom. ...  "

An Internet of Goods

From the Supply Chain Digest:

Supply Chain News: Here Comes the Internet of Goods

MAPI Research Report Predicts Major Manufacturing Transformation - to the Benefit of the US and the Rust Belt

SCDigest Editorial Staff
US manufacturing is set to see a game changing transformation, driven by digital technology – and it is coming just in time.

So writes Dr. Michael Mandel of the Progressive Policy Institute in a major report on web site of MAPI, the Manufacturing Alliance, a research organization focused on manufacturing and which is consistent provider of interesting insight. ... "

Wednesday, October 10, 2018

Maintenance via Sound Analysis

Not a new thing,  but useful as it could be used to augment human capabilities.  This had been suggested previously using neural, machine learning methods.

Sound Software for Fault Detection in Machinery 
 in CORDIS

A European Union-funded research project has developed software based on the human auditory system that can analyze sound to determine if industrial machinery requires maintenance. The Horizon2020 neuronSW team integrated advanced algorithms, machine learning, and big data analysis to mimic the human auditory cortex and enable early detection and prediction of breakdowns. Said SME NeuronSW Ltd.'s Jiri Cermak, "The technology leverages machine learning, the cloud, and the Internet of Things to deliver a detection service which emulates human intuition about sound." The neuronSW solution lets manufacturers perform intelligent audio diagnostics and monitor key pieces of machinery by the sounds they generate. Cermak said, "The integrated hardware and software platform automatically gather the sound of machines in real time and continuously assesses the equipment's health."  ... ' 

Tracking Plastics Use without Batteries

We looked at the problem of how pharmaceuticals were being used.   This would have solved one of the problems we examined.

3D-printed plastic objects can track their own use without any electronics  in TechnologyReview.

Objects such as pill bottles or prosthetics can now send information on how they are being used without the need for batteries.

How does it work? The researchers embed antennas in the objects in such a way that that are activated when the object is moved in a specific way—such as when a pill bottle opens or closes. The changes in how the two antennae backscattered signals that were transmitted to it created a tell-tale pattern that carried information about how and when it was being used.

Potential applications? The team at the University of Washington believes its devices could be perfect for assistive technology like prosthetics, “smart” pill bottles that remind patients to take their daily medications, and insulin pens. It would let you monitor exactly how people are using these devices, and it wouldn’t matter if they got wet, lost a Wi-Fi signal or ran out of battery.  ... " 

And more on this directly from the U of Washington. ...

GE Lighting, Google Home and Smart Homes

Lighting is the most obvious, simplest and obvious application in the smart home.  So this partnership could make for some real  infrastructure wins for Google,  easy of use is key.

GE and Google make using your smart bulbs as easy as saying, ‘Hey, Google’  in Digitaltrends

oogle has made a name for itself in the smart home market through Google Home, its smart artificial intelligence assistant. Now, Google plans to seize control of part of the smart lighting market through a partnership with GE. C by GE has introduced a new line of light bulbs that connect seamlessly to any of Google’s smart assistants without the need for a hub. Rather than go through a sometimes-complicated setup process, just screw in the light bulb and load up your Google Assistant app. It will automatically detect the bulb and let you pair it. From there, turning the bulb on or off, and adjusting the light level is as easy as saying, “Hey, Google.” ... " 

On Decision Responsibility

Yes, but also how decision links to further process and measured goals.

Articulation of Decision Responsibility    By Robin Hill in ACM

Remember the days when record-keeping trouble, such as an enormous and clearly erroneous bill for property taxes, was attributed to "computer error?" Our technological society fumbles the assignment of responsibility for program output. It can be seen easily in exaggerations like this, from a tech news digest: "Google's Artificial Intelligence (AI) has learned how to navigate like a human being." Oh, my. See the Nature article by the Google researchers [Google] for the accurate, cautious, descripton and assessment. The quote given cites an article in Fast Company, which states that "AI has spontaneously learned how to navigate to different places..." [Fast Company] Oh, dear.

But this is not the root of the problem. In the mass media, even on National Public Radio, I hear leads for stories about "machines that make biased decisions." Exaggeration has been overtaken by simple inaccuracy. We professionals in Tech often let this pass, apparently on the belief the public really understands that machines and algorithms have no such capacity as is normally connoted by the term "decision"; we think that the speakers are uttering our own trade shorthand. When we say that "the COMPAS system decides that offender B is more likely to commit another crime than is offender D" [ProPublica; paraphrase mine], it's short for "the factors selected, quantified, and prioritized in advance by the staff of the software company Northpointe assign a higher numeric risk to offender B than to offender D." When the Motley Fool website says "computers have been responsible for a handful of `flash crashes' in the stock market since 2010," it means that "reliance on programs that instantaneously implement someone's pre-determined threshholds for stock sale and purchase has been responsible... etc." [Motley Fool] .... "

Amazon Going Traditional Retail

Via Andreessen Horowitz by sonala16z

Amazon: Retailers Gonna Retail

The narrative of Amazon’s total disruption of traditional retail and ultimate dominance is well-known. What if many of the recent changes reflect not a path to domination, but Amazon’s journey as a more traditional retailer?

An annotated twitter thread. ....
Retail has always been generational. Retail success stories are a graveyard of iconic brand names. Just brutal in how they roll over every generation.

Here are a few from my own childhood that no longer exist. ... "

Mars is Going Digital

 Traditional candy company Mars moves Digital. Well put and in particular that it is about business process, Are we looking at the right problem?  Bot just wild experimentation.

Using Digital Tools to Move a Candy Company Into the Future  in NYTimes

Lauren Belomy, 32, is a digital transformation associate with Mars in New York City.

What do you do at work?

My job is to help move the company forward in a world that’s digital. We’re a traditional manufacturing company, and we need to be ready for the future.

I consider what opportunities the future will present, how we can best react and how we’ll make decisions about them.

I explore the ways in which we can take advantage of new technologies and tools, such as artificial intelligence; how we should experiment; and whether we are even looking at the right problems. Mars is based in McLean, Va., and I work remotely in Manhattan.

Did anyone do this before you?

It’s a new type of role, and one of my goals is to explain what’s possible in simple language so people understand.

Part of my work involves prototyping, such as growing peanut plants in a fish tank using digital automation — without human intervention. To do this, I worked with a few colleagues in Mount Olive, N.J., a unit that I’m part of, though I don’t work there all the time. We implemented an automated watering and fertilizing schedule to see how the plants would grow .....

Tuesday, October 09, 2018

Machinify to Monetize Data, Automate AI

First I heard of this, on the surface of it quite a useful claim, and starts from a data-based goal.    Where will it work?  Probably only in certain problem domains.   We spent much time looking at how to monetize data, think about it as an asset, and this could be one channel to consider.

Machinify raises $10 million to help businesses use AI to monetize data
By Megan Rose Dickey  @meganrosedickey in TechCrunch

Data is valuable — if you know how to access it and reap the insights from it. That’s where Machinify comes in. The artificial intelligence company just raised a $10 million Series A round led by Battery Ventures with participation from GV and Matrix Partners.

“Our core notion is that today, enterprises are collecting a ton of data,” Machinify  founder and CEO Prasanna Ganesan told TechCrunch. “But if you look at how many of them are successful in turning it into smarter decision-making to drive efficiency, very few companies are succeeding.”

With Machinify, enterprise customers feed the system raw data, specify what they’re trying to optimize for — whether that be revenue or some other goal — and then the machine figures out what to do from there. Based on past decisions, the machine can figure out the right thing to do, Ganesan said.

A good example of how companies use Machinify is in the healthcare space, where businesses are using the tool to increase the accuracy and speed with which they process claims. By doing so, these companies have been able to increase revenue and reduce costs. .... " 

Reading from their site at: https://www.machinify.com/home/  seems more a move to automate AI based on goals. ... "

NASA and Other Users of Hololens

Handsfree with reference to documentation and process.    Similar to our experiments.

NASA is using HoloLens AR headsets to build its new spacecraft faster

Lockheed Martin engineers wear the goggles to help them assemble the crew capsule Orion—without having to read thousands of pages of paper instructions.
by Erin Winick  in Technology Review

When you work at a factory that pumps out thousands of a single item, like iPhones or shoes, you quickly become an expert in the assembly process. But when you are making something like a spacecraft, that comfort level doesn’t come quite so easily.

“Just about every time, we are building something for the first time,” says Brian O’Connor, the vice president of manufacturing at Lockheed Martin.

Traditionally, aerospace organizations have replied upon thousand-page paper manuals to relay instructions to their workers. In recent years, firms like Boeing and Airbus have started experimenting with augmented reality, but it’s rarely progressed beyond the testing phase. At Lockheed, at least, that’s changing. The firm’s employees are now using AR to do their jobs every single day.  ... "

Using Email

Have read many of these kinds of email use hints, this one notes the need to understand the received tone, good point.  While some now believe that Email is over, its not.

Don’t send an email without doing these three things  in FastCompany By Stephanie Vozza

You probably don’t put much thought into crafting an email, which means it’s likely you are making some potentially embarrassing mistakes.   ... "

Amazon's Ad Business Booming

This completely surprised me, from CNBC. What does it mean for other Amazon Channels? Much more stats in the article.

" ... Amazon's ad business is booming. Some advertisers are moving more than half of the budget they normally spend with Google search to Amazon ads instead, amounting to hundreds of millions of dollars, according to execs at multiple media agencies. Some of these execs requested anonymity as they are not authorized to discuss their clients' expenditures in public.

Amazon's growing success could pose a rare threat to Google parent company Alphabet, which generated $95.4 billion in ad revenues last year, 86 percent of its total revenue. Google is the dominant digital advertising platform in the U.S., and will take in an estimated 37 percent of digital ad budgets in 2018. Although Alphabet does not disclose the breakdown of its ad revenue, most estimates believe the vast majority comes from search ads — approximately 83 percent in the year to date, according to research from eMarketer. .... "

Cisco Joins Blockchain Alliances

In a further indication of their seriousness in the space, Cisco has joined more BlockChain alliances.  Ecosystems are key and that is what Cisco is about.  They outline why they are doing this:

Why Cisco Has Joined Three More Blockchain Alliances   By Anoop Nannra in Cisco Blog

As we consider the complexity and vastness of the journey we’ve been on with blockchain, it’s become evident there’s still more work to be done at the ecosystem level. Cisco is a big believer in strategic partnerships in the blockchain space — that’s why we’re co-founding members of the Trusted IoT Alliance and Hyperledger.  Additionally, that’s why we have joined three more initiatives.

Blockchain is but a tool in a big tool chest of technologies, but it’s an especially important one because it can be a binding factor to drive transformational change. We’ve recently joined three more blockchain initiatives to help further advance research in a number of areas:

Facebook Hatches a Portal

Not so much an assistant, but multi person video chat.  Still an incomplete idea, the review says, but with hints it can be enhanced from the hardware.   No attempts at skills yet, but basic music too.  Not quite sure why Facebook is attempting this,  but it does complete the experience, for those that like Facebook, and for those that need family chat, why not.   As the name suggests, a portal.

Facebook tries its hand at hardware with Portal
The teleconferencing device comes in 10 and 15 inch models
By Brian Heater @bheater in TechCrunch

Facebook Portal

Portal is not Facebook’s Echo Show. Call it a case of convergent evolution, wherein two companies arrived at similar looking products after approaching hardware from different angles. The problem Facebook sought to solve is one of face to face communication. It’s an attempt to remove the device from the act of video chatting.

That Facebook, Amazon and Google’s smart display partners all ended up at a similar place is no coincidence, of course. Like those smart displays, the home teleconferencing device is essentially a propped up tablet. With Portal, however, the system takes two distinct form factors.

There’s the standard Portal, which looks quite a bit like Lenovo’s recently released Google Assistant Smart Display, and the more compelling Portal Plus. That larger model, with a 15 inch display (1920 x 1080) brings to mind recent enterprise attempts at telepresence robotics. The base is stationary here, but the display orientation can be swiveled into landscape or portrait mode.  ,,, " 

Monday, October 08, 2018

Linkedin Adds Glint

Nice to see additions added to the capabilities of Linkedin.   Have long thought there were lots of useful enterprise resources that could leveraged by the quantity and specifics of data here. What will Glint add?

LinkedIn acquires employee engagement platform Glint
By Ingrid Lunden @ingridlunden in TechCrunch

LinkedIn, the social network for the working world with close to 600 million users and now under the wing of Microsoft, has announced an acquisition as it continues to work on expanding the ways that people already on the platform use it. It has acquired Glint, a startup that provides employment engagement services for businesses and other organizations.

Terms of the deal are not being disclosed. For some context, Glint had raised nearly $80 million — including these rounds for $27 million and and $20 million in the last two years — was valued at around $220 million in its last round according to PitchBook. Investors included Bessemer Venture Partners, Norwest Venture Partners, Shasta Ventures and Meritech Capital Partners.

The news was announced both by LinkedIn and Glint itself in blog posts.

Daniel Shapero, VP of Talent Solutions at LinkedIn, said that the team from Glint will join LinkedIn and continue to work as a salient entity within it under current Glint CEO and founder Jim Barnett. ... " 

Whats an Inspector General?

My college roommate writes.

Who you gonna call? Inspectors General.  By John F. Sopko in The Hill

 In 1984, Bill Murray and Dan Aykroyd popularized that line – “Who you gonna call?” – from the theme to “Ghostbusters.” In 2018, however, “if there’s something weird/and it don’t look good,” the safest call may be to an inspector general. From allegations of FBI skullduggery, to accusations of profligate spending by cabinet offices, to theft and corruption surrounding the $126 billion spent on Afghanistan’s reconstruction, the inspector general community is increasingly in the news and relied upon for unvarnished, independent assessments on matters of grave national concern.

While many Americans, including some government officials, don’t know what an inspector general is, the concept is hardly new. Most Offices of Inspectors General were created by Congress in 1978, but the history of IGs goes back to 1778, when George Washington nominated Friedrich von Steuben to the newly created post of Inspector General of the Continental Army.

Von Steuben’s heirs – the 72 serving federal inspectors general – identify waste, fraud, and abuse in their respective agencies, and arrest those who steal from U.S. taxpayers.

This independent work is important – but it is also critical that inspectors general support Congress as it carries out its oversight obligations. Woodrow Wilson wrote that “it is the proper duty of a representative body to look diligently into every affair of government …the informing function of Congress should be preferred even to its legislative function.” .... " 

Startups aim to Disrupt Consumer goods

Startups shook up the sleepy razor market. What’s next? 

NEW YORK (AP) — What do you hate shopping for? Toothpaste? Sunscreen? The guys who founded Harry’s shaving club want to know.

The startup, which took on razor giants Gillette and Schick with its direct-to-consumer subscription model, has since expanded into traditional retail and launched a line of body care products. Armed with $112 million in new financing to develop new brands, the company now is investigating what other sleepy products might be ripe for disruption.

“It might be better products, a better experience getting the products or a brand that appeals to who they want to be as people,” said Jeff Raider, who recently took on the role of CEO of Harry’s Labs, overseeing the development of new brands.

There’s a reason why Harry’s investors are betting that reinventing the razor was no flash-in-the-pan idea. Insurgent brands are shaking up the way people buy everything from mattresses to prescription acne remedies, eating into the market share of big consumer product companies and leaving them scrambling to respond.

Eager venture capitalists, digital technology and social media make it easier for anyone with a good idea to enter the consumer goods market, according to a report on insurgent brands by Bain & Company, a management consulting firm. Contract manufacturing, which allows companies to outsource production and sometimes defray costs, also has made it simpler.

“The reality is that no category is immune to disruption,” the Bain & Company report said. ... " 

Escorting Customers Through Buying Journeys

Like the term 'escorting' here, the details of this view at the link.   More common for complex decisions.

Escorting Your Customers Through the Buying Journey
Ben Taylor  in CustomerThink

When your customer embarks on the path of making a purchase the road ahead is often long and complex. Information overload, a changing cast of stakeholders, and the pressure to deliver results makes overseeing a buying decision overwhelming.

Instead of a straight line, the buying journey is now cyclical, as show in the image below:

Amazon Brand Registry

Saw that Amazon was pushing this concept lately.  Also that my former employer, P&G is a participant.  What other kind of information patterns can be derived from such registrations?   Seems the data has other value once the architecture is set up.

Helping you protect your brand on Amazon

Amazon Brand Registry helps you protect your intellectual property and create an accurate and trusted experience for customers on Amazon.

Why Amazon Brand Registry?

Accurate brand representation
Once you enroll, Brand Registry gives you greater influence and control over your brand’s product listings on Amazon.

Powerful search tools

Amazon Brand Registry enables you to easily find content in different Amazon stores. Search for content using images, keywords, or a list of ASINs in bulk and report suspected violations through a simple, guided workflow.

Proactive brand protection

Our automated protections use information about your brand to proactively remove suspected infringing or inaccurate content. The more information you provide, the better Brand Registry can help you protect and improve your brand experience.

Learn more about the benefits of Brand Registry. .... 

 " ... P&G was very pleased to be invited to be an early adopter of Amazon Brand Registry. We believe that Amazon Brand Registry will continue to help as we collaborate with Amazon to protect our brands and our consumers online."....  '

Dynamic Programming Methods

Below was given earlier this week, Patrick Madden is developing an APP to analyze problems so you can determine what methods will work well.  Below the fold are some technical details about the App and its projected availability.    Combinatoric and dynamic programming a longtime interest of mine.  I plan to test this as it becomes available.  Technical.

Dynamic Programming for the Masses

Original Presentation at UC:  Slides.

Prof Patrick Madden, SUNY Bighamton.   Patrick.Madden@gmail.com


Splunk Used by Carnival Cruise

Been a while since I have looked at Splunk.  But here now, emphasizing data democratization:

Carnival gets data into all hands on deck with Splunk  By R. Danes in Siliconangle

If there’s one nugget among the heap of big data tools, tricks, and marketing fluff that businesses can actually use, it could be data democratization. This takes big data from a backroom concept to something business people can polish their product with day-in-day-out..... "

Sunday, October 07, 2018

Rethink Robotics Shuts Down

We followed this effort for some time,  notably the Baxter robot, but it may well depend on how you define collaborative robotics.  Where does the task begin and end?  How do you train humans to be usefully collaborative?   Should you wait for maximum robotic autonomy?    Below the starting excerpt, later in the piece,  founder Rodney Brooks is interviewed, which is insightful.  He suggests much has been learned and sales efforts may have been lacking.

Rethink Robotics, Pioneer of Collaborative Robots, Shuts Down

Rodney Brooks’s startup developed a new class of factory robots that could safely work alongside people. Then came the hardest part: selling them

By Erico Guizzo and Evan Ackerman in IEEE Spectrum

The Bad Boy of Robotics was in high spirits. Rodney Brooks had a lot to show us as we zigzagged through Rethink Robotics’ office in Boston on a June morning in 2012. That was the day Brooks introduced us to Baxter, a robot he said would transform manufacturing and “sell like hotcakes.”

So it was with a bit of sadness that we found out last night that Rethink is shutting down. After a long, 10-year run (the company was founded as Heartland Robotics, remember?) and US $150 million raised from investors, Rethink ran out of steam as sales fell short of the company’s goals. The news of the closing was first reported by the Robot Report yesterday. ... "

Quantum Computing Workplace

Technology continues to generate new jobs. The prime question remains application to real world problems. No doubt that such systems are being built.  A bit like when we saw neural networks become available in the 90s, and few uses were there.

U.S. Takes First Step Toward a Quantum Computing Workforce

MIT Technology Review    By Will Knight

In an effort to make the quantum computing industry more viable, the U.S. House of Representatives on Thursday passed legislation that would establish a federal program to advance research and training in quantum computing. Drafted with help from University of Maryland quantum physicist Christopher Monroe, the bill will release $1.275 billion in funding for centers of excellence that will train quantum engineers. To develop quantum computers that can handle real-world problems, Monroe says the U.S. needs engineers who understand quantum physics and the principles of computer engineering. Monroe says it is challenging for companies to find engineers to develop scalable systems, pointing to IonQ, the quantum computing startup that he co-founded. In the next five years, quantum computers will run calculations that could never have worked with conventional hardware, according to Monroe. However, these early systems will only be capable of certain types of computation, and determining how to use quantum systems will fall to quantum software engineers. Monroe says, "When we build them, they will be useful for something.".... "

Amazon Alexa Redesigns App, Interface

Good idea, the former design, which we used to implement our smart home,  was confusing.  We will be using the redesign to update a more multi room interaction.  Also new uses of better Ring camera integration for security.

Amazon’s Alexa app gets new interface to simplify use, control more devices  by Bruce Brown
@brucebrown   in DigitalTrends

Amazon is rolling out a new design for the Alexa app to help users manage smart home devices in multiple rooms. The new app has a more colorful design and should make it easier for users to configure and access their Alexa-compatible devices, Tech Crunch reported.

The app redesign is better described as a “rollout-in-process” because Amazon is releasing a series of Alexa App improvements, according to Amazon’s Help and Customer Service page.  The iOS Alexa app was released first. Amazon has not announced a date for the Android version, but the company hasn’t officially announced the new app yet, possibly because they’ll wait until the majority of new app features and versions for both major mobile OSs are ready.  ... " 

Alchemy.Codes for eCommerce Experience

Brought to my attention.  Via Procter & Gamble   Alchemy.codes  Much more at the link.  Note the mention of the Olay Skin Advisor, which I have written about here.  Recall P&G had previously tried startup style initiatives with Tremor, which utilizes consumer networks for brand engagement and efficient marketing.  Also relates to the Mr Clean and stain advisor work based on chatbot style AI systems.  All discussed here.

About Alchmy.codes:  (excerpts)

Our aim is to build disruptive technologies and business models that give our partners an unfair competitive advantage in the marketplace.

TRANSACTIONAL LEARNING EXPERIMENTS (TLES)

Thoughtful experimentation is in our DNA. We help you gauge new business opportunities through rapid prototyping and consumer feedback. Get to experimentation phase in days instead of months.

DIRECT-TO-CONSUMER 3.0 (DTC)
From strategy to execution, we provide everything that your brand needs to create a rich eCommerce experience. But we don’t stop at delivery - your Alchemy team will then obsess over growing your online business through constant experimentation, user testing and in-market iterations.

Our tried-and-tested approach enables us to provide strategic direction, data-informed designs, and thoughtful execution.

The voice of the consumer — ensures feature prioritization, roadmap management, and product success through critical KPIs (Key Performance Indicators).

Featured Work:    Olay Skin Advisor
Redesigned for excellence and rebuilt for performance, the new skin advisor has redefined the standard for brand engagement. We’re calling it, the gold standard.

Alchemy is reinventing the way that P&G brands connect with and deliver value to consumers.... "   
 i

Saturday, October 06, 2018

Manipulating P Values

More from Vincent Granville in DSC on P-Values and their misuse.  I recall the repeating experiments as a trick in an early statistical analysis course.  perhaps this is more common in other areas, but never saw this kind of misuse in engineering.  Useful read here.

Statistical Significance and p-Values Take Another Blow

I read an article this morning, about a top Cornell food researcher having 13 studies retracted, see here. It prompted me to write this blog. It is about data science charlatans and unethical researchers in Academia, destroying the value of p-values again, using a well known trick called p-hacking, to get published in top journals and get grant money or tenure. The issue is widespread, not just in academic circles, and make people question the validity of scientific methods. It fuels the fake "theories" of those have lost faith in science.

The trick consists of repeating an experiment sufficient many times, until the conclusions fit with your agenda. Or by being cherry-picking about the data you use, or even discarding observations deemed to have a negative impact on conclusions. Sometimes, causation and correlations are mixed up on purpose, or misleading charts are displayed. Sometimes, the author lacks statistical acumen.

Usually, these experiments are not reproducible. Even top journals sometimes accept these articles, due to:

Poor peer-review process

Incentives to publish sensational material

By contrast, research that is truly aimed at finding the truth, sometimes does not use p-values nor classical tests of hypotheses. For instance, my recent article comparing whether two types of distributions are identical, does not rely on these techniques. Also the theoretical answer is know, so I would be lying to myself by showing results that fit with my gut feelings or intuition. In some of my tests, I clearly state that my sample size is too small to make a conclusion. And the presentation style is simple so that non-experts can understand it. Finally, I share my data and all the computations. You can read that article here. I hope it will inspire those interested in doing sound analyses. ... "

Defining Cyber Battlespace

Interesting view.    Also to the extent it can be integrated with some measure, and prediction of risk.    Classic approaches for risk analysis calculates likely, statistical views in a context we are operating in.   The notion of a 'Battlespace', to me indicates, that I have an adversarial agent, which likely has overlapping goals.   Understanding threats and analyzing them is the first step to making something actionable.   Actionable in context also means the role of the agent (human or AI, or likely augmented human) that can respond.  Classification leads to better response.               

Defining Your Cyber Battlespace With Threat Intelligence
By Daniel Kropp on October 4, 2018

A Conceptual Framework for Threat Intelligence

An intelligence preparation of the battlefield (IPB) is a method for defining and understanding a specific operating environment in all battlespaces — a conceptual framework to better understand how to apply your threat intelligence. It has a been a staple tool to help military leadership make decisions, and when the same methodology is applied to the cyber realm, the benefits can be tremendous. By providing a broad overview for cybersecurity decision makers to aid in strategy development, it helps answer the basic question, “What is my current state of security?”

It can also provide a tactical understanding that helps decision makers implement more effective mitigating controls. Everyone in cybersecurity constantly seeks “actionable” threat intelligence, but it’s critical to realize that what is considered actionable depends on the role of the person asking the question. A well developed IPB provides a level of understanding that benefits everyone from the C-suite to the Tier 1 SOC analyst.  .... "

Honda Smart Intersection View

Clever, we are starting to see better collaboration between people and machines.  Here a good example.  Will kind of car augmentation delay complete self-driving?  Note this is now in test, and requires the intersection to become 'smart' as well by adding infrastructure.

Honda's smart intersection tells drivers what's around the corner
This intersection sees more than you or your car can.

By Roberto Baldwin, @strngwys in Engadget

Roles Driving AI Projects

In Forbes, this follows on some of my breadth points. Think of any technical roles being able to either be one of ...  these roles or influence them.

Top 3 Roles Driving A Majority Of AI Projects  in Cognitive World
in Forbes: Naveen Joshi Contributor   ....

We know that the ability of AI to drive organizational growth, innovate internal workflows, automate tedious and mundane jobs, enrich customer experience, and meet business goals is just incredible. Hence, in this era of business transformation, organizations are readily utilizing AI and its potential to partake in the technological revolution and innovation in business. Leveraging AI can either make or break an organization, depending completely on ‘how’ and ‘where’ the technology is used. However, leveraging the technology is secondary. Organizations should understand that the innovative technology, AI, can only deliver the expected benefits if the people associated with an AI project collaborate with the technology in the right way. The roles driving AI projects in any organization include business leaders, IT experts and customer service representatives. Each of these roles must perform their responsibility to the best fo their ability to drive an AI project to success.

Roles driving AI projects

Every organization aims to achieve successful AI implementation, yet AI will work faster, give optimum results, reduce operational costs, and enhance workforce productivity only if the employees within the organization are hard-working, talented, and skillful. Let’s check out the different roles associated with AI projects:   .... "

Friday, October 05, 2018

Rethinking Sales and Marketing

Have always been interested in how resources are delivered, and a big part of that is how the resources are organized and classified for useful application. 

Rethinking The Sales And Marketing Organization  By Dave Brock in Customerthink

We continue to organize our sales and marketing initiatives around what makes us more efficient or old views of how customers buy.

Classically, marketing’s focus is on creating interest and awareness, then driving demand.  The work toward MQLs, turning them over to sales, hopefully as SALs, saying “Good luck and godspeed!, we caught ’em, you skin ’em.”

Sales picks up the process, SDRs call to qualify the opportunity, they hand the lead to an account manager who gets more information, the customer is handed over to a pre-sales person for a demo, then someone else try to close them.

The overall marketing/sales assembly line takes customers through this linear process, all oriented to moving the customer through a buying decision.

Of course there are variants to this, there may be an account focus or orientation, there may be some sort of nurturing loop for customers that are not ready to buy, but as soon as we can we want to drop them into our marketing and sales assembly line.

Except our assembly line/linear customer engagement model doesn’t reflect how our customers buy.  Our customers aren’t engaged in a linear buying process, in fact, when we start mapping it, the process is very complex.  It’s a series of starts, stops, changes, reassessment, abandonment, restarting, stopping, going backwards, restarting, changing scope, adding new buying team members, more changes………

But we aren’t structured to be able to respond to the customer’s chaotic buying process.

How do we rethink things to have half a chance of intersecting the customer wherever they are at and trying to move them forward? .... "

Difficulties in Commonsense Reasoning

Abstract of good overview survey paper.

Commonsense Reasoning and Commonsense Knowledge in Artificial Intelligence   By Ernest Davis, Gary Marcus

Communications of the ACM, September 2015, Vol. 58 No. 9, Pages 92-103  10.1145/2701413

 ... discuss the shortcomings of AI systems and "Commonsense Reasoning and Commonsense Knowledge in Artificial Intelligence" (cacm.acm.org/magazines/2015/9/191169), their Review Article in the September 2015 Communications of the ACM.     https://vimeo.com/134965536

" ... To achieve human-level performance  in domains such as natural language processing, vision, and robotics, basic knowledge of the commonsense world—  time, space, physical interactions, people, and so on—will be necessary. 

Although a few forms of commonsense  reasoning, such as taxonomic reasoning  and temporal reasoning are well understood, progress has been slow. 

Extant techniques for implementing  commonsense include logical analysis,  handcrafting large knowledge bases,  Web mining, and crowdsourcing. Each of  these is valuable, but none by itself is a full solution. 

Intelligent machines need not replicate  human cognition directly, but a better  understanding of human commonsense  might be a good to start.  ... "

GE and BP with Machine Learning

Deep Machine Learning: GE And BP Will Connect Thousands Of Subsea Oil Wells To The Industrial Internet ... 

That magic is also the idea behind a new partnership between BP and GE’s Intelligent Platforms unit. Next year, BP will use GE’s Predix software platform to connect 650 wells to the Industrial Internet. If all goes according to plan, the companies will expand the scope to 4,000 BP subsea wells around the world. “We will help them to get connected, get insights and get optimized,” says Kate Johnson, CEO of GE Intelligent Platforms Software. ... " 

Free Open Source Training Data

Always looking for good open source data, especially for testing models.  William Vorhies of of DSC talks this.  Sources and links.  Good examples:

Lots of Free Open Source Datasets to Make Your AI Better     Posted by William Vorhies  

Summary:  There are several approaches to reducing the cost of training data for AI, one of which is to get it for free.  Here are some excellent sources.

Recently we wrote that training data (not just data in general) is the new oil.  It’s the difficulty and expense of acquiring labeled training data that causes many deep learning projects to be abandoned. 
It also matters a great deal just how good you want your new deep learning app to be.  A 2016 study by Goodfellow, Bengio and Courville concluded you could get ‘acceptable’ performance with about 5,000 labeled examples per category BUT it would take 10 Million labeled examples per category to “match or exceed human performance”. 

There are a number of technologies coming up through research now that promise more accurate auto labeling to make creating training data less costly and time consuming.  Snorkel from the Stanford Dawn Project is one we covered recently.  This area is getting a lot of research attention.  ... " 

Thursday, October 04, 2018

The Threat Intelligence Handbook

Just reading.   Well done.  See recordedfuture.com      Get the free book here.

The Threat Intelligence Handbook   (108 pages)

A Practical Guide for Security Teams to
Unlocking the Power of Intelligence

Edited by Chris Pace
Foreword by Dr. Christopher Ahlberg

It’s easy to find descriptions of what threat intelligence is. But it’s harder to learn how to use it to truly make your organization safe from cybercriminals. How can threat intelligence strengthen all the teams in a cybersecurity organization?

This book answers this question. It reviews the kinds of threat intelligence that are useful to security teams and how each team can use that intelligence to solve problems and address challenges. It discusses how security analysts in the real world use threat intelligence to decide what alerts to investigate (or ignore), what incidents to escalate, and what vulnerabilities to patch. It examines how information collected outside of the enterprise can help model risks more accurately and prevent fraud.

We invite you to learn about how threat intelligence can help everyone in cybersecurity anticipate problems, respond faster to attacks, and make better decisions on how to reduce risk. ... "

Ford Looks for Common Car Language

Makes lots of sense.

Ford asks for a common language for self-driving cars
It wants every car to signal its intentions the same way.

Jon Fingas, @jonfingas in Engadget

Ford has been developing its own means for self-driving cars to communicate their intent, but it also knows that fragmentation could be a huge problem in the autonomous driving world. How are you supposed to know what cars are doing when each one has a different visual cue? Accordingly, Ford has issued a memorandum of understanding that asks the industry to create a signalling standard for cars capable of at least SAE Level 4 automation (that is, they can handle all driving tasks under some conditions). It hopes the exchange of standards will lead to a common language that human drivers and pedestrians will understand, regardless of where they live or how tech-savvy they may be.... " 

Wal-Mart Sets Standards for SC Blockchains

An indication that big enterprise are serious, a useful example.

Walmart Sets Blockchain Requirement for Greens Suppliers   By CGT Staff  

Walmart will require all suppliers of leafy green vegetables to its Sam’s Club and Walmart stores to upload data to its blockchain by September 2019. The announcement comes after the retail giant spent more than a year working with IBM on a food safety blockchain initiative.

Called the Food Trust Solution, the new technology digitizes the produce supply chain process using the IBM Blockchain Platform. This makes the supply chain traceable and transparent. Each node on the blockchain could represent an entity that has handled the food, making it easier, faster and more precise to trace an infected produce batch back to a particular farm.

“Our customers deserve a more transparent supply chain,” said Frank Yiannas, Walmart’s vice president of food safety. “We felt the one step up, one step back model of food traceability was outdated. This technology-supported move will greatly benefit customers and transform the food system, benefitting all stakeholders.”


Traceability is particularly important in produce. Retailers generally use multiple suppliers per category to cope with fluctuations in demand, crop size and seasonality. They also use myriad shippers, distributors and other “middlemen.”

Without blockchain, it takes about seven days to trace the source of food. Blockchain can do it within 2.2 seconds, significantly reducing the likelihood that infected food will reach consumers.

Currently, most supply chains are bogged down in manual processes. This makes it difficult and time consuming to track the origins of products like the E. coli-infected romaine lettuce that affected the industry last spring.

The IBM Blockchain Platform was created to help companies build, govern and run blockchain networks, according to Bridget van Kralingen, IBM’s senior VP for global industries, platforms and blockchain. It uses Hyperledger Fabric, an open source digital ledger technology, and runs on the IBM Cloud.

Suppliers do not have to be blockchain experts to adopt the technology, according to Walmart spokesperson Molly Blakeman. They simply need an internet connection and know how to upload data to the blockchain application. Currently, suppliers keep records using everything from basic paper and Excel spreadsheets to sophisticated ERP systems.

“IBM will offer an onboarding system that orients users with the service easily,” added Blakeman. “Essentially, suppliers will need a smart device and Internet to participate.”

In addition to the blockchain mandate, Walmart is requiring that suppliers adhere to the Global Food Safety Initiative, an internationally recognized set of food safety standards.

According to a report from Global Market Insights, the blockchain market is expected to surpass $16 billion by 2024. Digital currency and retail supply chain (particularly fresh food and apparel) are among the biggest areas of application.   .... " 

Omohundro on Costly Signalling

Interesting thoughts, can this be implemented in a behaviorally predictive system?

What Scientific Term or Concept Ought to be more Widely Known?
By Steve Omohundro   Scientist, Self-Aware Systems; Co-founder, Center for Complex Systems Research

Costly Signalling

If something doesn't make sense, your go-to hypothesis should be "costly signalling." The core idea is more than a century old, but new wrinkles deserve wider exposure. Veblen's "conspicuous consumption" explained why people lit their cigars with $100 bills as a costly signal of their wealth. Later economists showed that a signal of a hidden trait becomes reliable if the cost of faking it is more than the expected gain. For example, Spence showed that college degrees (even in irrelevant subjects) can reliably signal good future employees because they are too costly for bad employees to obtain.

Darwin said, "The sight of a feather in a peacock's tail, whenever I gaze at it, makes me sick!" because he couldn't see its adaptive benefit. It makes perfect sense as a costly signal, however, because the peacock has to be quite fit to survive with a tail like that! Why do strong gazelles waste time and energy "stotting" (jumping vertically) when they see a cheetah? That is a costly signal of their strength and the cheetahs chase after other gazelles. Biologists came to accept the idea only in 1990 and now apply it to signalling between parents and offspring, predator and prey, males and females, siblings, and many other relationships.

Technology is just getting on the bandwagon. The integrity of the cryptocurrency "bitcoin" is maintained by bitcoin "miners" who get paid in bitcoin. The primary deception risk is "sybil attacks," where a single participant pretends to be many miners in an attempt to subvert the network's integrity. Bitcoin counters this by requiring miners to solve costly cryptographic puzzles in order to add blocks to the blockchain. Bitcoin mining currently burns up a gigawatt of electricity, which is about a billion dollars a year at US rates. Venezuela is in economic turmoil and some starving citizens are resorting to breaking into zoos to eat the animals. At the same time, enterprising Venezuelan bitcoin miners are using the cheap electricity there to earn $1200 per day. Notice the strangeness of this: By proving they have uselessly burned up precious resources, they cause another country to send them food!

As a grad student in Berkeley, I used to wonder why a preacher would often preach on the main plaza. Every time he would be harassed by a large crowd and I never saw him gain any converts. Costly signalling explains that preaching to that audience was a much better signal of his faith and commitment than would preaching to a more receptive audience. In fact, the very antagonism of his audience increased the cost and therefore the reliability of his signal.

A similar idea is playing out today in social media. Scott Alexander points out that the animal rights group PETA is much better known than the related group Vegan Outreach. PETA makes outrageous statements and performs outrageous acts which generate a lot of antagonism and are therefore costly signals. They have thrown red paint on women wearing furs and offered to pay Detroit water bills for families who agree to stop eating meat. They currently have a campaign to investigate the Australian who punched a kangaroo to rescue his dog. Members who promote ambiguous or controversial positions signal their commitment to their cause in a way that more generally accepted positions would not. For example, if they had a campaign to prevent the torture of kittens, everyone would agree and members wouldn't have a strong signal of their commitment to animal rights. ... 

This connects to meme propagation in an interesting way. Memes that everyone agree with typically don't spread very far because they don't signal anything about the sender. Nobody is tweeting that "2+2=4." But controversial memes make a statement. They cause people with an opposing view to respond with opposing memes. As CGP Grey beautifully explained, opposing memes synergistically help each other to spread. They also create a cost for the senders in the form of antagonistic pushback from believers in the opposing meme. But from the view of costly signalling, this is good! If you have enemies attacking you for your beliefs, you better demonstrate your belief and commitment by spreading them even more! Both sides get this boost of reliable signalling and are motivated to intensify meme wars. ... " 

Wal-Mart Tests Arc de Triomphe Model for Pickup

The 'hand-off to the customer' part of pickup continues to be examined.

Wal-mart expands test of giant automated grocery kiosk    by Tom Ryan in Retailwire with expert commentary.

After a test in Oklahoma City last summer, Walmart is opening a second grocery pickup kiosk at a location in Sherman, Texas that enables customer to pick up orders without having to interact with a store associate.

Under the set up:

- Orders are placed by customers who shop online or through their mobile browser at walmart.com/grocery;
- Walmart associates inside the store fill the orders;
- The orders are organized in bins and placed in the massive kiosk located in the parking lot, which is    equipped with refrigerators and freezers for perishable goods.
- Customers pull up to the kiosk building, walk up to an interface station and scan the barcode they received with their “order ready” e-mail. The kiosk retrieves the order, delivering it to the customer in a process that takes a minute or less.  .... "

Demystifying AI for Execs

Yet another piece to de-mystify AI for execs.  Used to be what I did.  Not bad, a useful description that is worth reading.  Good that it is mentioned that it is about how people and machines will ultimately work together.  But I have a number of problems with it.    First, hardly a mention of decisions and process.    And ultimately that is what it is all about.  If we don't know how people and decisions interact, we don't know much.  Also no mention of goals, which is another key component.  What are we trying to do?  Risk too is not mentioned.  What happens if the result is wrong, or poorly applied to real decisions?  Not much about operations. 

Also, more fundamentally, the article talks mostly about components of  'AI' like image recognition, or classification or clustering.   There is no doubt we can do these much better than we every could.  Great progress.  But AI is much more than that.  These things are newer, faster, sharper tools.  But without the process of taking data, knowledge and process and embedding them in business,  you don't get much.  Please make that clear when you talk to execs. .

" ... Demystifying AI and machine learning for executives

In this interview, Tamim Saleh cuts through the hype around artificial intelligence with guidance for executives about where and how to employ AI in their businesses. ... " 

Wednesday, October 03, 2018

Network of Brains

Voluntarily, I hope.  If this works trying to think the implications.

Technical reference:    Ref: arxiv.org/abs/1809.08632: BrainNet: A Multi-Person Brain-to-Brain Interface for Direct Collaboration Between Brains

The first “social network” of brains lets three people transmit thoughts to each other’s heads

BrainNet allows collaborative problem-solving using direct brain-to-brain communication.   by Emerging Technology from the arXiv    in TechnologyReview

The ability to send thoughts directly to another person’s brain is the stuff of science fiction. At least, it used to be.

These tools include electroencephalograms (EEGs) that record electrical activity in the brain and transcranial magnetic stimulation (TMS), which can transmit information into the brain.

In 2015, Andrea Stocco and his colleagues at the University of Washington in Seattle used this gear to connect two people via a brain-to-brain interface. The people then played a 20 questions–type game.

An obvious next step is to allow several people to join such a conversation, and today Stocco and his colleagues announced they have achieved this using a world-first brain-to-brain network. The network, which they call BrainNet, allows a small group to play a collaborative Tetris-like game. “Our results raise the possibility of future brain-to-brain interfaces that enable cooperative problem-solving by humans using a ‘social network’ of connected brains,” they say. .... "

Wearable Goal Cockpit Concept

Displays based on predicted goals are generated and produced

An innovative wearable cockpit in design by BAE could lighten the load on both fighter planes and their pilots.

The wearable cockpit concept would rely on modern technology, such as gesture and eye-tracking technology, that would eliminate many of the physical components typically manipulated by the pilot. In their place, the pilot would wear a VR-enabled helmet that would project the readouts and controls in an efficient and reconfigurable way. 

According to the team, however, the standout feature of the system is its ability to infer the pilot’s goal based on where they are looking, using AI to help anticipate the necessary tools. ... "

Deep Learning without Labels for Apache Spark

With a quite instructive visual example with the Snow Leopard images.

Machine Learning Blog

Deep Learning Without Labels    by ML Blog Team 

Announcing new open source contributions to the Apache Spark community for creating deep, distributed, object detectors – without a single human-generated label

This post is authored by members of the Microsoft ML for Apache Spark Team – Mark Hamilton, Minsoo Thigpen, Abhiram Eswaran, Ari Green, Courtney Cochrane, Janhavi Suresh Mahajan, Karthik Rajendran, Sudarshan Raghunathan, and Anand Raman. 

In today’s day and age, if data is the new oil, labelled data is the new gold.

Here at Microsoft, we often spend a lot of our time thinking about “Big Data” issues, because these are the easiest to solve with deep learning. However, we often overlook the much more ubiquitous and difficult problems that have little to no data to train with. In this work we will show how, even without any data, one can create an object detector for almost anything found on the web. This effectively bypasses the costly and resource intensive processes of curating datasets and hiring human labelers, allowing you to jump directly to intelligent models for classification and object detection completely in sillico.

We apply this technique to help monitor and protect the endangered population of snow leopards.

This week at the Spark + AI Summit in Europe, we are excited to share with the community, the following exciting additions to the Microsoft ML for Apache Spark Library that make this workflow easy to replicate at massive scale using Apache Spark and Azure Databricks:

Bing on Spark: Makes it easier to build applications on Spark using Bing search.
LIME on Spark: Makes it easier to deeply understand the output of Convolutional Neural Networks (CNN) models trained using SparkML.

High-performance Spark Serving: Innovations that enable ultra-fast, low latency serving using Spark.

We illustrate how to use these capabilities using the Snow Leopard Conservation use case, where machine learning is a key ingredient towards building powerful image classification models for identifying snow leopards from images.  ... "

Tennenbaum Institute

Just reviewed Ga Tech's Tennenbaum Institute.  Had not for some time.  Especially about process and decision in the enterprise.  Worth a look:

The Tennenbaum Institute at Georgia Tech helps businesses and organizations use data to inform better decision making.

Our research and education addresses a rich mixture of concepts, principles, models, methods, and tools applicable to a wide range of enterprise domains, with particular emphasis on healthcare delivery and global manufacturing. Our work in these two domains is enhanced by strong partnerships with leading companies and agencies, as well as thought leaders from many organizations. We believe that this diversity is our greatest strength in addressing the complex organizational problems the world faces today.

The Tennenbaum Institute was founded in 2004, enabled by the generosity of Georgia Tech alumnus, Michael Tennenbaum.  Our Research Approach: 

The Tennenbaum Institute has established a structured approach for understanding the transformation of complex enterprise systems. This approach includes four fundamental steps as outlined below.  .... 

Notable too their book: Work, Workflow and Information Systems.   Has probably since. been updated.

Tuesday, October 02, 2018

Diffbot : Foundation of Knowledge Graphs

A massive public database and graph.  How might it be integrated with private data? 

In Datanami, a description of the Diffbot Graph
The Graph That Knows the World   By Alex Woodie

Somewhere in a data center in Fremont, California, exists a large computer cluster that’s hoovering up every piece of data it can find from the Web and using machine learning algorithms to find connections among them. It’s arguably the largest known graph database in existence, encompassing 10 billion entities and 10 trillion edges.

No, it’s not some secret government project to catalog the world’s information. In fact, the graph was created and is run by a private company called Diffbot, and in fact you can get access to it for as little as $300 per month.

You can’t accuse Mike Tung, the founder and CEO of Diffbot, of thinking small, or beating around the bush for that matter. During an interview last week, he got right to the point. “The purpose of our company,” he tells Datanami, “is to build the first comprehensive map of all human knowledge.”

That might sound like a crazy thing to do, in 2018, a quarter century after the Web went mainstream, after the first dot-com crash, the rise of Web 2.0, the emergence of e-commerce 3.0, and the forthcoming industry 4.0 wave that’s projected to shake it all lose again. Haven’t we done this already? And isn’t that what Google and Wikipedia are for?


Diffbot CEO and founder Mike Tung graduated from Stanford University with a master’s degree in AI

Not according to Tung, who started work on the Diffbot graph while at Stanford University in 2008 and then started the Diffbot company in 2011. While it’s true that Google and Wikipedia are creating large knowledge graphs, they’re not as useful as one might think, Tung says.

“Our knowledge base is not only larger, deeper and more accurate [than Google’s and Wikipedia’s] but it’s accessible and more useful,” Tung says. “We hope that this is the first step in creating a future where…you have almost infinite access to knowledge.”

AI Crawlers

Tung says that what makes Diffbot unique, apart from its size and public nature, is how it’s assembled. While Google and Wikipedia rely largely on human labor to curate the information that goes into their graphs – and Facebook relies on its 2 billion users to create its knowledge graph —  the Diffbot graph is created automatically  — autonomously, really — through a variety of machine learning techniques, including computer vision, natural language processing (NLP), and others.

The Diffbot knowledge base currently has 10 billion vertices, which correspond to entities, including people, places and things. Connecting those 10 billion entities are 10 trillion edges, which are facts that can be searched through an API or DQL, the SQL-like Diffbot Query Language. ... "