Robot Record Sales

Expect to see robotics in many new places.

World Record Sales for Robots as Sector Reaches $16.5 Billion in Investment   in ZDNet  By Greg Nichols

The International Federation of Robotics' (IFR) World Robotics Report found that 422,000 robotic units were shipped globally in 2018, an increase of 6% compared to 2017. Among the takeaways from the report was an increase of 23% in annual installations of collaborative robots from 2017 to 2018. China continues to be the world's largest industrial robot market, accounting for 36% of total units installed. Robot installations in the U.S. reached about 40,300 units in 2018, 22% more than the year prior. The report highlighted the growing use of robotics in sectors like construction, mining, and healthcare, as technology developers respond to labor crunches following strong global economies. Said Junji Tsuda, president of IFR, "We saw a dynamic performance in 2018 with a new sales record, even as the main customers for robots—the automotive and electrical-electronics industry—had a difficult year." ... '

The AI Work of the Future Report

Having been seeing the AI hype of late, and getting questions from colleagues and clients as to that tt really means.  This piece from MIT, pointed to by O'Reilly, is refreshing in that addresses what still cannot be done, needs to be addressed.   Their "Work of the Future Report: Shaping Technologies and Futures".

I am a big proponent and optimist on the topic, but still think we need to know what the unsolved challenges still are.   And make plans as to what we need to do to solve them.   Both by the scientists and by the business decision makers and process inventors.   Do read this report.

Tracking Drugs with Blockchain

Another example of blockchain use for secure tracking and thus tracing.    The regulation implied is not to specifically use blockchain, but to ensure secure tracking.

How pharma will soon use blockchain to track your drugs in Computerworld

Under regulatory pressure, a large number of pharmaceutical manufacturers, shippers and wholesalers are adopting blockchain to track and trace prescription drugs.
By Lucas Mearian  .... "

Snorkel for Building Data for ML

This was new to me.  But handling and selecting the data is the most important aspect of machine learning projects.   In a recent project it included over 75% of the resource effort. And likely to be much more of the ongoing maintenance effort.  Worth a good look.

Introducing Snorkel
How this Tiny Project Solves One of the Major Problems in Real World Machine Learning Solutions

By Jesus Rodriguez Towards data Science.

Building high quality training datasets is one of the most difficult challenges of machine learning solutions in the real world. Disciplines like deep learning have helped us to build more accurate models but, to do so, they require vastly larger volumes of training data. Now, saying that effective machine learning requires a lot of training data is like saying that “you need a lot of money to be rich”. It’s true, but it doesn’t make it less painful to get there. In many of the machine learning projects we work on at Invector Labs, our customers spend significant more time collecting and labeling training dataset than building machine learning models. Last year, we came across a small project created by artificial intelligence(AI) researchers from Stanford University that provides a programming model for the creation of training datasets. Ever since, Snorkel has become a regular component of our machine learning implementations.  .... " 

Company Building Brands to Sell only on Amazon

A direction we may see more of.

How one company is building brands to sell only on Amazon    By Cale Guthrie Weissman in ModernRetail

Many digital brands are allergic to Amazon. Some, however, welcome it with open arms.

A growing number of companies pushing products are being built specifically for Amazon — forgoing Google and Facebook for customer acquisitions and instead going all in on the e-commerce platform. Driving this is Innovation Department, an agency that through its media assets and email data has been able to build a suite of brands that are Amazon only. By leveraging millions of email addresses, the company has been able to funnel traffic to its products on Amazon, which has shuttled them to the top spots of product search.

Innovation Department is based in New York, and has a few businesses under its umbrella. One is a piece of software for brands to collaborate and manage consumer email data called DojoMojo. Another is a media company called Valyrian Media, which produces a series of weekly newsletters on a range of topics — from food to fashion to technology. Alex Song, the company’s founder and CEO, said the emails are similar to The Skimm; he uses a roster of 25 freelance writers to keep it going. Through those two engines, Innovation Department claims it has built up an audience of over 1.5 million subscribers.  .... " 

Optimal Neural Architecture

Thoughtful and useful piece.    Though I don't seen how this is necessarily universally optimal, which is usually a broad claim.  Link to full and technical paper below.

How to Construct the Optimal Neural Architecture for Your Machine Learning Task

By Adrian de Wynter
Alexa Alexa research Alexa science

The first step in training a neural network to solve a problem is usually the selection of an architecture: a specification of the number of computational nodes in the network and the connections between them. Architectural decisions are generally based on historical precedent, intuition, and plenty of trial and error.

In a theoretical paper I presented last week at the 28th International Conference on Artificial Neural Networks in Munich, I show that the arbitrary selection of a neural architecture is unlikely to provide the best solution to a given machine learning problem, regardless of the learning algorithm used, the architecture selected, or the tuning of training parameters such as batch size or learning rate.

Rather, my paper suggests, we should use computational methods to generate neural architectures tailored to specific problems. Only by considering a vast space of possibilities can we identify an architecture that comes with theoretical guarantees on the accuracy of its computations.

In fact, the paper is more general than that. Its results don’t just apply to neural networks. They apply to any computational model, provided that it’s Turing equivalent, meaning that it can compute any function that the standard computational model — the Turing machine — can.

To be more specific, we must introduce the function approximation problem. This is a common mathematical formulation of what machine learning actually does: given a function (i.e., your model) and a set of samples, you search through the parameters of the function so that it approximates the outputs of a target function (i.e., the distribution of your data). ..... " 

Material Holes Create Amazing Properties

An interesting discovery of the use of  'holes'.  Consider all the advantages we have gotten from material science.

Researchers catalog defects that give 2-D materials amazing properties
Theoretical analysis distinguishes observed “holes” from the huge list of hypothetically possible ones.  David L. Chandler | MIT News Office

Amid the frenzy of worldwide research on atomically thin materials like graphene, there is one area that has eluded any systematic analysis — even though this information could be crucial to a host of potential applications, including desalination, DNA sequencing, and devices for quantum communications and computation systems.

That missing information has to do with the kinds of minuscule defects, or “holes,” that form in these 2-D sheets when some atoms are missing from the material’s crystal lattice.

Now that problem has been solved by researchers at MIT, who have produced a catalog of the exact sizes and shapes of holes that would most likely be observed (as opposed to the many more that are theoretically possible) when a given number of atoms is removed from the atomic lattice. The results are described in the journal Nature Materials in a paper by graduate student Ananth Govind Rajan, professors of chemical engineering Daniel Blankschtein and Michael Strano, and four others at MIT, Lockheed Martin Space, and Oxford University.  ... "

Self-Flying Cargo Drones

Emergence of such capabilities will change transport.

Bell's New, Self-Flying Cargo Drone Hauls a Heavy Load in Wired
The all-electric APT 70 can tote up to 70 pounds, cruise at 75 mph, and cover 35 miles with a fully charged battery.  ... "

Podcast Interview with Hilary Mason on GigaOM

Another AI practitioner talks about the advances and future of AI:

Voices in AI – Bonus: A Conversation with Hilary Mason   By Byron Reese

On this Episode of Voices in AI features Byron speaking with Hilary Mason, an acclaimed data and research scientist, about the mechanics and philosophy behind designing and building AI.

Listen to this episode or read the full transcript at www.VoicesinAI.com

Byron Reese: This is Voices in AI, brought to you by Gigaom and I am Byron Reese. Today, our guest is Hilary Mason. She is the GM of Machine Learning at Cloudera, and the founder and CEO of Fast Forward Labs, and the Data Scientist in residence at Accel Partners, and a member of the Board of Directors at the Anita Borg Institute for Women in Technology, and the co-founder of hackNY.org. That’s as far down as it would let me read in her LinkedIn profile, but I’ve a feeling if I’d clicked that ‘More’ button, there would be a lot more.

Welcome to the show, amazing Hilary Mason!

Hilary Mason: Thank you very much. Thank you for having me.

I always like to start with the question I ask everybody because I’ve never had the same answer twice and – I’m going to change it up: why is it so hard to define what intelligence is? And are we going to build computers that actually are intelligent, or they can only emulate intelligence, or are those two things the exact same thing?

This a fun way to get started! I think it’s difficult to define intelligence because it’s not always clear what we want out of the definition. Are we looking for something that distinguishes human intelligence from other forms of intelligence? There’s that joke that’s kind of a little bit too true that goes around in the community that AI, or artificial intelligence, is whatever computers can’t do today. Where we keep moving the bar, just so that we can feel like there’s something that is still uniquely within the bounds of human thought.

Let’s move to the second part of your discussion which is really asking, ‘Can computers ever be indistinguishable from human thought?’ I think it’s really useful to put a timeframe on that thought experiment and to say that in the short term, ‘no.’ I do love science fiction, though, and I do believe that it is worth dreaming about and working towards a world in which we could create intelligences that are indistinguishable from human intelligences. Though I actually, personally, think that it is more likely we will build computational systems to augment and extend human intelligence. For example, I don’t know about you but my memory is horrible. I’m routinely absentminded. I do use technology to augment my capabilities there, and I would love to have it more integrated into my own self and my intelligence. ..... " 

Advances in AI Earthquake Prediction

We attended some early neural net applications meeting where this was proposed, and added some of our own thoughts.   Nice to see this is evolving.   Are there shaking patterns in the earth that reliably predict earthquakes?    Thinking it is likely yes, but enough for the prediction and likely magnitude and location of major events?   I think yes too

AI Helps Seismologists Predict Earthquakes  in Wired
Machine learning is bringing seismologists closer to an elusive goal: forecasting quakes well before they strike. .... 

Artificial Intelligence Takes On Earthquake Prediction in QuantaMag

After successfully predicting laboratory earthquakes, a team of geophysicists has applied a machine learning algorithm to quakes in the Pacific Northwest.

In May of last year, after a 13-month slumber, the ground beneath Washington’s Puget Sound rumbled to life. The quake began more than 20 miles below the Olympic mountains and, over the course of a few weeks, drifted northwest, reaching Canada’s Vancouver Island. It then briefly reversed course, migrating back across the U.S. border before going silent again. All told, the monthlong earthquake likely released enough energy to register as a magnitude 6. By the time it was done, the southern tip of Vancouver Island had been thrust a centimeter or so closer to the Pacific Ocean.

Because the quake was so spread out in time and space, however, it’s likely that no one felt it. These kinds of phantom earthquakes, which occur deeper underground than conventional, fast earthquakes, are known as “slow slips.” They occur roughly once a year in the Pacific Northwest, along a stretch of fault where the Juan de Fuca plate is slowly wedging itself beneath the North American plate. More than a dozen slow slips have been detected by the region’s sprawling network of seismic stations since 2003.  And for the past year and a half, these events have been the focus of a new effort at earthquake prediction by the geophysicist Paul Johnson.    ..... " 

Alexa Skills for Productivity

Still, I think not good enough to really make me have a standard device on my desk at work.   What can be done to really make it essential?

Review: 18 Alexa skills for productivity, collaboration and more in Computerworld

You can use Amazon’s voice-activated Alexa assistant to send Slack messages, texts, and emails; add items to to-do lists; and more. But do Alexa skills for business users really save you time and effort?
         
 By James A. Martin

Earlier this year, Amazon announced it had sold more than 100 million Alexa devices. Along with the consumer market, Amazon is also pushing Alexa into offices via Alexa for Business, which enables developers to create skills exclusively for internal users at their companies via APIs and other tools.

But can Alexa’s off-the-shelf skills truly make enterprise users more productive? Will they make collaboration easier? To find out, I tested 18 Alexa productivity and collaboration skills that are available to everyone but potentially useful for business professionals. All of these skills are free, although some are associated with paid or freemium services, as noted.  .... "

Learning and Revealing Private Data

Been looking at past articles of the Berkeley AI Group, and found an interesting aspect of data privacy examined.  Can a neural network, while being trained,  inadvertently learn and thus reveal pieces of data that happen to be in the presented data?  So say if a credit card number was in the data, could it later reveal that if the trained model was examined?  And what could you do about it?    Nicely done, largely non technical piece.

Evaluating and Testing Unintended Memorization in Neural Networks
By Nicholas Carlini    Aug 13, 2019

It is important whenever designing new technologies to ask “how will this affect people’s privacy?” This topic is especially important with regard to machine learning, where machine learning models are often trained on sensitive user data and then released to the public. For example, in the last few years we have seen models trained on users’ private emails, text messages, and medical records.

This article covers two aspects of our upcoming USENIX Security paper that investigates to what extent neural networks memorize rare and unique aspects of their training data.  (The paper's abstract provides a further descriptive overview)

Specifically, we quantitatively study to what extent following problem actually occurs in practice:

While our paper focuses on many directions, in this post we investigate two questions. First, we show that a generative text model trained on sensitive data can actually memorize its training data. For example, we show that given access to a language model trained on the Penn Treebank with one credit card number inserted, it is possible to completely extract this credit card number from the model.

Second, we develop an approach to quantify this memorization. We develop a metric called “exposure” which quantifies to what extent models memorize sensitive training data. This allows us to generate plots, like the following. We train many models, and compute their perplexity (i.e., how useful the model is) and exposure (i.e., how much it memorized training data). Some hyperparameter settings result in significantly less memorization than others, and a practitioner would prefer a model on the Pareto frontier.    .... "

5G Coverage for IOT

Was recently asked to give an opinion of 5G use in the Cincinnati area for potential IOT applications,  with mobility implications,  and was pointed to this map.  Which can be used US country wide.  This particular map gives you only AT&T 3G to 5G.   It implies it is frequently updated.  You can click on the map for many locations in the US, zoom in, etc.   Useful for early analyses of applications.  etc.   Please pass along pointers to any other resources of this type.

Information Latency Study for DOD

I  suggest that there are important latency conditions in many parts of large networked systems.  For example in supply chains it can greatly change costs, effective responses, contract and goal compliance, risk analysis,  decision design integration, etc.    Information latency is always considered in such systems, but often not carefully enough.  Latency is a key kind of metadata, and should be included in a 'knowledge graph' to represent a problem in both its statement and in any automated approaches being designed.   - Franz

Research Team to Study Information Latency With $7.5M DOD Grant
By Virginia Polytechnic Institute and State University
 Virginia Tech researchers Walid Saad, Jeffrey Reed, and Thomas Hou

Information latency is a measure of how quickly or slowly networked devices transmit information. When the information being transmitted is for the military, understanding latency may be the deciding factor in the outcome of warfare.

That's one of the reasons the U.S. Department of Defense has now tapped the expertise of an interdisciplinary research team led by Virginia Tech to study latency and information freshness in military Internet of Things systems with a $7.5 million, five-year Multidisciplinary University Research Initiative (MURI) grant.

The goal is to develop a novel foundational framework for guaranteeing low latency and information freshness in military networked systems, such as the Internet of Things, using a cutting-edge concept known as multimode age of information, which tightly ties in information latency with the dynamic networked military system.

The project will fundamentally define this new concept of information latency and provide a suite of tools to optimize multimode age of information in massive-scale military networked systems.

"Despite much progress being made in the study of military communications, the basic science for tracking, control, and optimization of information latency is yet to be developed," says principal investigator Jeffrey Reed, Willis G. Worcester Professor of Electrical and Computer Engineering in the College of Engineering at Virginia Tech. "In fact, a fundamental knowledge of information latency is crucial for our military to maintain information superiority on the battlefield."  ..... " 

(see more at link above)

Do we Know How the Brain Works?

Have had  conversations of late with people who have said:  Look at neural nets they are modeled after brains.    But the answer is still, no we don't.  And we are still not close.   Artificial neural models are very different even from the way we think we know how biological neurons work.    Not to say the artificial models are not useful,  but its not what your brain actual does.  How much closer are we getting?  See the Neuralink approach, mentioned below, to understand the challenges.  Will we know?  I am always optimistic.

Will It Ever Be Possible to Understand the Human Brain?
Despite technical breakthroughs like Elon Musk’s Neuralink, scientists still have no reliable model of how the brain actually works

By Brian Bergstein in Medium ... 

Structured Signals for Model Training

Technical but interesting point about how to add structured knowledge into otherwise non transparent networks.  Examining further.

Posted by Da-Cheng Juan (Senior Software Engineer) and Sujith Ravi (Senior Staff Research Scientist)

We are excited to introduce  Neural Structured Learning in TensorFlow, an easy-to-use framework that both novice and advanced developers can use for training neural networks with structured signals. Neural Structured Learning (NSL) can be applied to construct accurate and robust models for vision, language understanding, and prediction in general.

Neutral structured learning framework

Many machine learning tasks benefit from using structured data which contains rich relational information among the samples. For example, modeling citation networks, Knowledge Graph inference and reasoning on linguistic structure of sentences, and learning molecular fingerprints all require a model to learn from structured inputs, as opposed to just individual samples. These structures can be explicitly given (e.g., as a graph), or implicitly inferred (e.g., as an adversarial example). Leveraging structured signals during training allows developers to achieve higher model accuracy, particularly when the amount of labeled data is relatively small. Training with structured signals also leads to more robust models. These techniques have been widely used in Google for improving model performance, such as learning image semantic embedding.

Neural Structured Learning (NSL) is an open source framework for training deep neural networks with structured signals. It implements Neural Graph Learning, which enables developers to train neural networks using graphs. The graphs can come from multiple sources such as Knowledge graphs, medical records, genomic data or multimodal relations (e.g., image-text pairs). NSL also generalizes to Adversarial Learning where the structure between input examples is dynamically constructed using adversarial perturbation.  ... " 

See also:  https://www.datanami.com/2019/09/04/google-adds-structured-signals-to-model-training/

See also:  https://venturebeat.com/2019/09/03/google-launches-tensorflow-machine-learning-framework-for-graphical-data/ 

Sensing and AR/VR

Good to see AR/VR linked strongly to sensing capabilities.     As is suggested this is the way we construct models of the word.  Whether they be virtual or real life.   It also allows us to link data to those worlds and drive to better solutions via analytics or AI.

3 Questions: Why sensing, why now, what next?  in MIT News
By Brian Anthony, co-leader of SENSE.nano, discusses sensing for augmented and virtual reality and for advanced manufacturing.

MIT.nano

Sensors are everywhere today, from our homes and vehicles to medical devices, smart phones, and other useful tech. More and more, sensors help detect our interactions with the environment around us — and shape our understanding of the world.

SENSE.nano is an MIT.nano Center of Excellence, with a focus on sensors, sensing systems, and sensing technologies. The 2019 SENSE.nano Symposium, taking place on Sept. 30 at MIT, will dive deep into the impact of sensors on two topics: sensing for augmented and virtual reality (AR/VR) and sensing for advanced manufacturing. 

MIT Principal Research Scientist Brian W. Anthony is the associate director of MIT.nano and faculty director of the Industry Immersion Program in Mechanical Engineering. He weighs in on why sensing is ubiquitous and how advancements in sensing technologies are linked to the challenges and opportunities of big data.

Q: What do you see as the next frontier for sensing as it relates to augmented and virtual reality?

A: Sensors are an enabling technology for AR/VR. When you slip on a VR headset and enter an immersive environment, sensors map your movements and gestures to create a convincing virtual experience.

But sensors have a role beyond the headset. When we're interacting with the real world we're constrained by our own senses — seeing, hearing, touching, and feeling. But imagine sensors providing data within AR/VR to enhance your understanding of the physical environment, such as allowing you to see air currents, thermal gradients, or the electricity flowing through wires superimposed on top of the real physical structure. That's not something you could do any place else other than a virtual environment.    .... " 

Apple Shows Interest in Blockchain Tech

The fact that Apple is following this is significant.  Apple Pay at least could have future implementations to consider.    Comments below.

Cryptocurrency Has ‘Long-Term Potential,’ Says Apple Exec
Apple is “watching cryptocurrency,” according to an executive at the tech giant.

Apple Pay vice president Jennifer Bailey, talking to CNN at a private event in San Francisco, said “We think it’s interesting. We think it has interesting long-term potential.”Apple is “watching cryptocurrency,” according to an executive at the tech giant.  Apple Pay vice president Jennifer Bailey, talking to CNN at a private event in San Francisco, said “We think it’s interesting. We think it has interesting long-term potential.”

Bailey did not elucidate about the possible uses of the technology Apple might pursue. She had been taking about the future of payments at the CNN event.  With Facebook planning to launch its Libra stablecoin next year, it would be surprising indeed if Apple were not watching crypto. But, Bailey’s comments may come as confirmation that more might be going behind the scenes at Apple’s Cupertino HQ.

In February, Apple submitted a filing with the Securities and Exchange Commission (SEC) that contained rare details about the computing giant’s interest in blockchain tech.  .... "

Bailey did not elucidate about the possible uses of the technology Apple might pursue. She had been taking about the future of payments at the CNN event.

With Facebook planning to launch its Libra stablecoin next year, it would be surprising indeed if Apple were not watching crypto. But, Bailey’s comments may come as confirmation that more might be going behind the scenes at Apple’s Cupertino HQ.

In February, Apple submitted a filing with the Securities and Exchange Commission (SEC) that contained rare details about the computing giant’s interest in blockchain tech.   ..... " 

Google Quantum Supremacy?

Quite a tease here.   Have they really reached this goal?  And what was the nature and form of the problem?  See much more below.    And at the link.

Google researchers have reportedly achieved “quantum supremacy”
Google's quantum computer  in MIT Tech Review

The news: According to a report in the Financial Times, (oops, the site is walled) a team of researchers from Google led by John Martinis have demonstrated quantum supremacy for the first time. This is the point at which a quantum computer is shown to be capable of performing a task that’s beyond the reach of even the most powerful conventional supercomputer. The claim appeared in a paper that was posted on a NASA website, but the publication was then taken down. Google did not respond to a request for comment from MIT Technology Review.   ... " 

Apple Overton Leading to Code Automation?

Increasingly moving towards automating many aspects of coding.   In fact robot assistants that 'observe' the coding process could readily insure that secure, robust and repeatable methods were used when building AI systems.    They could also make sure that the most important methods were shared, maintained and updated as new research dictated. 

On the data side, that the data was properly selected, prepared and delivered with needed metadata to support explainable results.  That's why I am not a believer in just training everyone in low level coding.  People are not good at these skills.    Train them in problem solving supported by prefabricated AI systems and results visualization methods, because ultimately the classic methods will be built, solved, updated and delivered by automation.

Apple ‘Overton’: Automating Low-Code Machine Learning     By Nick Kolakowski

Apple has struggled in recent years to establish a robust artificial intelligence (A.I.) practice. This partially stems from the company’s ironclad privacy policies—it’s more difficult to analyze datasets for insights when internal rules prevent the company from using every piece of user data it can vacuum up. Nonetheless, Apple’s newest projects show that it’s powering ahead anyway—including one platform that, if it’s ever released, could change how you use A.I. and machine learning (ML).

(It’s worth remembering how, in a 2015 speech, Apple CEO Tim Cook accused tech giants such as Facebook and Google of “gobbling up everything they can learn about you and trying to monetize it,” which he framed as “wrong.” It seems unlikely that Apple’s stance on data and privacy will change during Cook’s tenure.)

According to a just-released paper with the dry-but-mysteriously-compelling title “Overton: A Data System for Monitoring and Improving Machine Learned Products,” a group of Apple researchers describe their work on a machine-learning platform (named—you guessed it—“Overton”) designed to “support engineers in building, monitoring, and improving production machine learning systems.”   ......... '

Abstract of paper mentioned above:    https://arxiv.org/pdf/1909.05372.pdf   (technical)

 ... We describe a system called Overton, whose main design goal is to support engineers in building, monitoring, and improving production machine learning systems. Key challenges engineers face are monitoring fine-grained quality,diagnosing errors in sophisticated applications, and handling contradictory or incomplete supervision data. Overton automates the life cycle of model construction, deployment, and monitoring by providing a set of novel high-level,declarative abstractions. Overton’s vision is to shift developers to these higher-level tasks instead of lower-level machine learning tasks. In fact, using Overton, engineers can build deep-learning-based applications without writing any codein frameworks like TensorFlow. For over a year, Overton has been used in production to support multiple applications in both near-real-time applications and back-of-house processing. In that time, Overton-based applications have answered billions of queries in multiple languages and processed trillions of records reducing errors 1.7 − 2.9× versus  production systems. .... "