Sils Maria Grison Switzerland

A window out

Every year has its moments, and 2016 was no exception. Various significant shifts occurred, including changes in the political landscape in the United States, United Kingdom, and Turkey. And the horrors of war, civil strife, terrorism and an underlying global tension have been constantly fed into our digital lives from the comfort of our screens.

As we consume the aggregated algorithmic social network feeds, each customized to ensure we get what we want to digest, we are choreographed into a more divisive world.

Information is power. This year, the pollsters, news agencies, and pundits got caught out with two big votes, and so many predictions seemed off.

Our landscape of information has entered a level of Orwellian curation, and what is news, fact, or reality seems dictated by emotion and perspectives constructed from our own curated news feeds. They are rarely factual. “Post Truth” – Oxford English Dictionary Names ‘Post-Truth’ Word of the Year by Jon Blistein is the word that defines these moments and a shift to a new narrative.

For many of us, this Orwellian curation has us struggling to distinguish fact from fiction. The level of sophistication of not only the algorithms but how these are manipulated to shift thinking is the new power. In schools, we are being told by various studies that our students capacity for media and information literacy is weak. (Students Have ‘Dismaying’ Inability To Tell Fake News From Real, Study Finds by Camila Domonoske ). When you consider we as adults struggle with this landscape, it is no surprise that our students struggle too.

In a world of algorithms where the sophisticated digital curation of social media, news, blogs, and video feeds can be manipulated to match an individual’s perspective, the challenges we face as educators are immense. This manipulation, shared in this sobering article “ Google, democracy and the truth about internet search by Carole Cadwalladr“, highlights the complexity of being truly media literate. The prevalence of third party curation in social media feeds during elections highlighted in this article “Macedonia’s fake news industry sets sights on Europe by: Andrew Byrne” emphasis the challenges we all face in understanding what is “real” news.

To be complacent is short-sighted in a school setting. There is a tendency with school professional development to not explicitly address the digital reality that engulfs our lives as an essential part of our professional learning. Information and Media literacy are what frame our own democratic values: choice, perspective, empathy, resilience, and critical thinking. If we as educators are going to assign students critical thinking tasks and ask them to engage with media and information while juggling screen time in a complex digital landscape, we cannot be passive bystanders.

As school leaders, we need to re-frame our engagement with the role of digital life in professional development. Together, we need to understand the complexity and impact of algorithmic information flows on our devices.

We also need dedicated spaces for this professional learning. We must learn how to mentor information flows, authenticate media, source perspectives, and understand the pedagogic impact of a curated news. We must approach this with patience and empathy, and allow everyone to build an understanding of the digital flows we live by, tapping into the talent of our librarians and digital coaches as guides. We must take advantage of the frameworks available to us (e.g: #1 or #2) and use them ourselves, as a point of reference for a pedagogic consensus on how to mentor our school community.

The paradigm shift asks us to look at Digital Intelligence as a core intelligence. As defined by http://www.projectdq.org: “- the sum of social, emotional, and cognitive abilities essential to digital life.” and shared out in the World Economic Forum article: “8 digital life skills all children need – and a plan for teaching them“.

Digital Intelligence needs to be woven into the curriculum. We do this on a daily basis with all other aspects of the curriculum. Let us do it with Digital Intelligence. Re-structure the focus and content to explicitly encompass screen-time management, privacy management, cyber security management, digital footprints, and digital identity; use these to make authentic connections based on our experiences. Then, reflect on our digital habits, likes, tensions, questions and understandings to create activities to share. In this process, we should hope to find comfort in being honest with our own vulnerabilities. We can then use this life-learning to support our students’ understanding of digital intelligence.

Being explicit about implementing Digital Intelligence in faculty professional learning ensures this is an essential part of our educators professional growth. Working together, as adults learners, we need to harness the complexity of the choreographed digital world. By ensuring this is in our professional learning landscape, we are then empowered to share our digital intelligence to students. It is the only way to counter an Orwellian curation of information in a “post truth” world.

John @ beyonddigital.org

a wonderful resource by Joyce Valenza to consider: Truth, truthiness, triangulation: A news literacy toolkit for a “post-truth” world

Welcome to the “GAFA” (Google, Apple, Facebook, Amazon) world. The”GAFA” world is where most of humanity’s internet users and consumers work, communicate, socialize, learn, entertain themselves, and share, in services provided by one, two, three or all four of these companies: the “GAFA” grids.

We have become comfortable with “GAFA’” being part of our lives in multiple venues, and as a result, schools, educators, students and parents are investing significant amounts of monies into “GAFA”. It is an essential component of our ability to function at school and at home, and the collective convenience and seamless experience of “GAFA” intoxicates us.

In Terry Heick’s (@TeachThought) thoughtful article “How Google Impacts The Way Students Think”, he highlights how learners working in a Google ecosystem develop an appetite for a black and white information age. The expectation? Immediate answers, 24/7. The convenience of this immediacy creates an illusion of thinking, but actually disengages the user from deep critical thinking. It does this by simplifying the process of gathering information and giving the impression it is all connected.

In order to have a constant infusion of innovation and creativity, “GAFA” also hungers for start-up companies. By absorbing these companies, they are able to facilitate the pollination of ideas, products and services and enrich their ability to generate more seamless methods of connectivity. In this way “GAFA’s” largeness and versatility is engrained in all aspects of our lives

This innovation also provides “GAFA” with opportunities to tie our lives closer together with multiple platforms and venues in a frictionless environment. Examples of this reach are Amazon’s cloud service, which hosts large architectures of company websites, services, and databases, including the CIA’s; Google moving into the home with Nest and pursuing the development of artificial intelligence (Dark Blue Labs and Vision Factory); Apple’s acquisition of Affectiva, a company that specializes in emotion recognition, and investments in health apps and services; and Facebook’s expansion into virtual reality. Making its services ubiquitous, as with the “free wi-fi-with-check-in ”in hotels and small businesses. Its purchase of “Whatsapp” is another example of how a “GAFA” company spent billions on an innovative service.

The algorithms provide a treasure trove of information with which to understand our behavior, habits, aspirations and desires. In Raffi Khatchadourian’s article “We Know How You Feel”, we are reminded that the hunger for data is tied to a hunger for emotional interactions. In Shelley Podolyn’s New York Times article, “If an Algorithm Wrote This, How Would You Even Know?”, she highlights the level of sophistication of writing algorithms generating news articles and books. In tandem, the growth of “The Internet of Eyes“ in objects we interact with, as part of the “ Internet of Things.” brings about a new dynamic to data mining. It is a reminder that many of these algorithms being designed within “GAFA” play an almost non-negotiable role in our lives.

Many schools believe that their curriculum’s should allow for authentic connections to the world around them. What about “GAFA”? Should we as learners, guides, mentors, and facilitators highlight “GAFA”? Is this important? Should its presence be considered in our learning outcomes? To ignore “GAFA” is to create a disconnect with present changes that are reshaping all of our lives. It sidelines a reality that is the future. What does “GAFA” mean, to us, our schools, community and educational institutions? Schools have a responsibility to ensure this is part of the curricular discourse. We need to construct learning moments and scaffold time to pause, reflect, understand, explain and critically think about what it is to live in a “GAFA” world.

If personal privacy, independent thought, critical thinking, differentiation, balanced perspectives, mindfulness and our capacity to be unique are in our school’s mission, we need to address what it means to be curated by “GAFA”. Will we not lose an important aspect of humanity, if we continue to ignore “GAFA”?

John@beyonddigital

P.S: Next time you are at a Starbucks drinking your coffee remember that the free wifi is a “GAFA” gift!

When we consider the components of these infrastructures, networks stand out. After all, without connectivity, none of the other parts of the infrastructure matter much.  Moreover, along with data transmission requirements, there are requirements for networking. One of those is network analysis, which has become more complex and demanding in this new age and which is not easily accomplished.

The requirements for network analysis include:

1. Capturing data that reach your networks at high speed, in a variety of data formats (streaming data) through various network protocols (NetFlow, PCAP, IPFIX, sFlow, GeoIP, BGP, SNMP and others)
2. Storing such captured data in data storage in real time
3. Analyzing such data to take appropriate action in real or near real time

The very rudimentary diagram below depicts these.

Figure 1: Oversimplified Kentik Architecture (diagram by the author)

I totally agree with Nelson’s comment, and I wondered whether a company existed that contributes to satisfying such demanding requirements for network analysis. I had an opportunity to talk to Alex Henthorn-Iwane late last year; I had interviewed him when he was with a previous company. His new company, Kentik, sounded interesting, and I recently got to talk to him and Jim Frey, VP, Product, at Kentik Technologies, Inc.

Jim Frey

As usual, I will summarize our discussion at a level deeper than a press release but appropriate for laymen.

Target Markets

According to Frey, their products and services are useful to cloud companies like Amazon and Microsoft, ISPs/telecom companies like AT&T and Verizon, and e-commerce companies like eBay. I think data center service providers, such as Equinix, could benefit from them. Actually, Kentik colocates at an Equinix data center and is in talks with them for adoption of their services.

Differentiations

I added my own to Frey’s, below.

1.    SaaS

Frey said providing services with SaaS is one of their differentiations. If you can set up their services easily, and you do not need to worry about placing any other appliances on your premises but can receive results, it would be best.

2.    Multitenancy

Another one is multitenancy. Multitenancy security could mean two different things: logical and physical data separations. Logical separation means each user’s data segments are placed on the same physical storage, and separation is done via software. Physical separation means each user’s data segments are placed on separate physical storage. Kentik implements logical separation now but plans to implement physical later. For those who are really worried about data security, their services can be hosted on-premise at each user’s site.

3.    Investments and management team

Their war chest is stuffed with about $15M from several investors. In addition, the management team consists of people in the relevant fields, such as CDN.

4.    Technologies

In addition to Frey’s points, the following, in my opinion, are noteworthy for this demanding digital age and differentiators:

1.      It can capture streaming data of various kinds and speed.

2.      It can store such data in a database in a real time.

3.      It can apply analysis quickly.

The current packet capture market requires 10 to 40 gigabits speed but will soon require 100 gigabits. Kentik provides adequate support for such high speeds. Points 2 and 3 are requirements for streaming data store and analysis, which is a trend in data analysis these days. I asked Frey for some details of the three components.

His answer was that all three were developed by Kentik from the ground up. The streaming data management and analysis include such technologies as the Lambda architecture by Nathan Marz and Apache Spark. There are a few more, like Storm, Flink, and Samza. Kentik does not use any of those, said Frey. In the past, Avi Freedman, CEO, Kentik, contributed to a Quora question, “What kinds of applications is Apache Spark not suitable for?” His answer was “Superfast semi-random access to superlarge persistent data stores.” I wonder what “superfast” and “superlarge persistent data stores” are, though. Kentik filed patents on these technologies but did not elaborate on them, as they are still in the patent process.

Data Center

Kentik colocates at an Equinix data center in Ashburn, Virginia. Incidentally, it is well known that if you start a data center, you want to start with one in Virginia. I thought this was quite a good move on their part. These days, it is becoming common knowledge that unless your power requirements are larger than several tens of MW, it does not make sense to own your data centers. In addition, I think that Equinix could use their services, but Frey declined to comment on that, other than to say that they are in discussions with them.

From another point of view, teaming up with Equinix is a really good idea. I think Equinix is disrupting the Internet market because users and their ISPs do not have to worry about peering too much, because Equinix now has direct connection services to major clouds, including Amazon, Microsoft, Google, and Softlayer (IBM). More and more people and companies are dealing with clouds. On top of that, Equinix has its data centers in strategic locations all around the world. By colocating with them, Kentik deploys their services worldwide without major efforts.

Analytics

One such analysis is the detection of DDOS. Frey said their analysis is along the lines of a well-known detection method. I could spend multiple blogs on this alone, but I won’t go into detail here. He said their customers want to know how to mitigate DDOS attacks, and they are working on it as well.

One interesting analysis is to optimize peering. As indicated above, the Internet market is changing. Bill Norton’s book is a good reference for understanding how the current Internet is implemented, with abundant peering information. More analytics can be developed by applying machine-learning technologies.

Summary

Because of the lack of time, I could not get into the details of the implementation, which attracts most of my interest. I want to find out more about their streaming data technologies, which are said (by its CEO) to be better than Spark. After all, IBM created Spark Technology Center because it invests in that technology so much.

Finally, Figure 1 is expanded into the following Figure 2.

Machine Learning (ML)

Hype

AI includes a wide variety of subfields, but recently the terms ML and/or deep learning (DL) have been used interchangeably with AI. In our interview, Dr. Kunio Matsui mentioned that a boom for ML comes every 15 years. Back around the year 2000, ML got a lot of attention. Then in 2015, there was another boom. I used the past tense because it is 2016 now. Matsui referred to Gartner’s Hype Cycle 2015. He pointed out that ML was no longer at the peak of inflated expectations. I checked it carefully and found ML is slightly off the peak. Matsui said that depending upon the use/area/application, the boom could last another 12 months or longer. But he said that it was inevitable to have another AI winter.

Source: Gartner 2015 Hype Cycle for Emerging Technologies August, 2015

Suitable Areas of Applications

Is ML a panacea for each and every area of problems? How does ML work?

Because I asked Matsui to keep his answers at a layman’s level, they were easy to follow intuitively. He described how ML worked, using a simple example. Suppose we have many beads and are to put each bead into one of a certain number of boxes according to some predefined criteria. If the number of boxes is small, it is easy to put each bead into the appropriate box. But if there are a huge number of boxes, it is not that easy, and sometimes impossible, to do so. Classifying emails into spam and not spam is relatively easy because there are only two boxes (spam or not spam). Matsui said that for NLP, there are far many more boxes than we have spam detection for, and it is usually hard to apply ML to that.

In the case of spam detection, performance has improved dramatically in the past 15 years. He said this was because of the availability of a large number of examples, on the order of hundreds of millions. Of course, the increase of computing power is another factor in better performance.

ML Bottom Line

He summarized three conditions in which ML works well:

1. Previous experience data exist

2. Some well-defined rules exist
3. The scope is well defined

I draw the conclusion that ML is not necessarily a panacea for every problem. Learning in ML still needs to be programmed prior to execution. On the other hand, humans can tackle something they have never encountered before with innovation and intuition, although previous education and experience may help. In other words, if we prepare everything (including the scope of the data sources to use and specific models of how to learn), ML can do very well and may actually do much better than humans. It is very important to understand which areas ML can and cannot be effectively applied to.

I asked Matsui about DL, which some consider to be ML on steroids. He said DL is a subfield of ML and has more power than normal ML. But even so, the bottom line is the same. Matsui said that NLP usually has far too many boxes, and it is very hard to apply ML effectively to NLP. It is very important to combine both rules and ML/DL for NLP.

Future

Some researchers and business leaders have expressed their concerns over the danger that AI could take over the world and harm humans. This may be attributed to the fear that strong AI/artificial general intelligence and the 2045 Singularity may become reality.

I asked Matsui if he worries about this problem. He was very clear on this subject and did not think AI would ever surpass human intelligence. I asked him twice if he really meant it. The answer was a resounding NO. In any event, I do not have to worry about this in my lifetime. In 2045, I may be around, but I am not sure that my mind will be sharp enough to understand what’s going on around me.

In the end, he listed three researchers (Japanese nationals) in the area of NLP who will lead this research area:

• Professor Kentaro Inui of Tohoku University
• Professor Sadao Kurohashi of Kyoto University
• Professor Satoshi Sekine of New York University

Summary

Dr. Matsui was very nice to answer my layman’s questions, which may have sounded silly to an expert like him. I often think that those who know the most can explain hard concepts most simply. That was exactly the case.

AI has been researched for many years and its scope is very wide. Anyone can write an article after reading five to ten articles on the subject. But it is very hard to grasp a succinct idea of what it is, where it stands now, and where it is going. I was very fortunate to be able to speak with Dr. Matsui. Based on our conversation, I conclude the following:

• ML can be applied to many areas, as long as the aforementioned three conditions hold. We have not exhaustively looked into its applicability yet. There are many vertical areas and subareas for consideration. The current AI/ML/DL boom will continue for some time until this exercise is completed.
• Those who successfully apply this will be strong proponents of AI, while those who cannot will downplay AI, leading to the third AI winter.
• The combination of well-defined rules and “brute force” (ML/DL) will allow further progress in NLP and other areas. When this is pushed until no more progress can be made, a new approach to AI will be considered and further advances can be made.
• I am not sure if the Biological Neural Network advocated by Jeff Hawkins will be the next stage of research, but it is certainly a good read, and I highly recommend your spending your time on it. The blog is not highly technical. Rather, it is written for the average Joe. Hawkins listed traditional AI (expert system), neural network (NN, including ML/DL), and biological neural network, which he advocates.

Finally, I would like to thank Mr. Hitoshi Matsumoto, Executive Advisor, Fujitsu Laboratories of America, Inc., for introducing me to Dr. Matsui.

To save space, I won’t define each term here. You can easily look it up by following the link:

• AI = Artificial Intelligence
• NLP = Natural Language Processing
• ML = Machine Learning
• DL = Deep Learning
• ANN = Artificial Neural Network

I warn you that this blog is not meant for NLP or AI experts but for the interested layman.

On a recent trip to Japan, I had a chance to talk with an expert in the field of NLP and artificial intelligence at Nifty Corporation, a subsidiary of Fujitsu, and whose business is providing online service in Japan. The expert’s name is Dr. Kunio Matsui and he is executive manager at Nifty. In the past, he was with Fujitsu’s research labs specializing in NLP.

Dr. Kunio Matsui

Initially, my motive was to find out how hard it is to Japanize a computer system. More specifically, I was curious why it was taking a long time for IBM Watson to Japanize its system. However, as we talked about this subject, our discussion drifted to cover many aspects of artificial intelligence, which was very, very useful for my understanding of the subject. AI is now resurgent and the market is confused, although AI is far from being a new field. It is hard to tell where AI stands now, where it is going, what we can do with the current AI, and what cannot be done. Also, in my view, the terms ML and AI are often used interchangeably, although ML is a subfield of AI and there are many more subfields in it. It is very important to keep your cool when analyzing AI to understand what it is, where it is now, and where it is going.

NLP

Architecture

I wanted to know how NLP interfaces with the AI engine. I prepared two different versions of the interface, as shown in Figure 1, and showed them to Matsui. Real systems are not this simple, but remember that I am approaching everything from a layman’s point of view. If the architecture is like the one on the left side of Figure 1, with a clearly defined interface (in yellow in the figure), all we have to do is to rip the English NLP and replace it with the Japanese one. If this is the case, it seems trivial to support multiple languages, including two-byte languages like Japanese.

In reality, things are not that easy. According to Matsui (he used my example and he did not actually use this figure), a system is architected like the one on the right side of Figure 1. Instead of the well-defined interface shown on the left side, NLP is usually tightly integrated with the engine. When we want to support a different language, it cannot be easily replaced. If we rip an NLP for one language and replace it with one for another language, it cannot fit into the points for the former languages. The new NLP needs more complex integration and takes expertise in the new language and efforts to adjust the integration. This explains why IBM is taking time to Japanize Watson.

 

Figure 1: How NLP is integrated with the AI engine. The left one has a clean interface between NLP and the engine. The right one is an architecture that tightly integrates NLP with the engine.

Japanization

I was wondering when a system can accurately be described as Japanized. Matsui, not knowing how IBM does so, speculated as follows. First of all, it is important to have knowledge or cortex in Japanese. All of the original Watson’s knowledge and dictionaries are in English and derived from many places, including Wikipedia. Matsui thought IBM was using the Japanese Wikipedia to Japanize Watson. Second, Q&As in Japanese have been categorized. This means that different types of questions are classified  in certain categories, and their answers can be determined according to categories retrieved. With these two, Watson can perform Q&A in Japanese. So in essence, the Japanese version of Watson is not that much different from DeepQA, which was a major component of the English version of Watson, victor of Jeopardy in 2011.

Present and Future of NLP

Although our conversation concerned the Japanese language, it is also applicable to other languages, like English. The current NLP can process natural languages well in terms of syntax and grammar. When it comes to semantics, it struggles except for  some easy cases. This is mainly because natural languages like Japanese and English are ambiguous. I mean that the same word may mean different things in different contexts. The English word spring can mean one of the four seasons or a curved metal. The only way to understand the true meaning of the sentence using spring is to understand the sentences before and after it. This is called semantics analysis. Matsui concluded that most natural languages suffer from this semantics analysis.

Matsui said that NLP should be approached by two different methods. One is to develop a set of rules applicable to each language. Even though each language is ambiguous, certain things can be well-defined by rules. The rest should be taken care of by “brute force” (this is not his word but the author’s). This “brute force” is ML/DL. I think this is similar to an empirical study when you cannot establish a theory of something’s behavior. This approach is very programmatic. In other words, this is an engineering approach rather than a purely scientific one. Being a former engineer himself, this author is in total agreement with this approach. Matsui also said that this is the current trend in the NLP research community.

Continued to Part 2

Disruption has already happened IBM: Source: IBM http://www.ibmforentrepreneurs.com

Disruption for many people generates discomfort, shakes the status quo and breaks routines. We all have an emotional response to this process: for some hesitation, doubt, confusion, fear, anger, bewilderment, and for others excitement, rejuvenation, inspiration, motivation and energizing or a combination of the above.

The Digital Disruption Has Already Happened” image on twitter challenged my own thinking, and as I unpacked this with a group of students we all had to take time to unravel what this meant to us. After a healthy discussion we came to a common understanding that each of these companies generated a “disruption” armed with ideas and models that completely reshape the economic contract of the business world. The disruption challenged a set of expectations, routines and structures, thus redefining what it means to provide a particular service. In the process, the relationship between worker, employer, customer and their social contracts was also redefined. As these models of disruption become part of our economic ecosystem, a whole new paradigm faces us. A good example of this process is featured in this article “The Last Kodak Moment?” (Economist)

As the students and I further discussed what this means to us here in the classroom, we started realizing a distinct disconnect between the objectives and outcomes of our learning in school and the reality that this image represents. There are two worlds working in very different realms with no clear bridge creating a meaningful authentic connection between them. As one student shared after seeing this video in class, “mmm I find it odd that we are not learning how to make these things in school, or understand how they work or maybe fix them.”

Yes, the world around us is changing rapidly, very rapidly, we know this, we are aware of this, we state this, and are impacted by this daily. However, our role is often that of the passive consumer, unconsciously sucked in by the addictive seamless convenience of the services these disruptions deliver. In our passivity and growing dependence we seem willing to sideline a more critical reflection of what this disruption is doing to us. The engines of this disruption: creativity, machine learning, algorithms, and innovation are driving the ecosystems which are quickly becoming non negotiables with our work and social flows. These disruptions are inevitable and not bad or good, they are part of today’s economic narrative that impacts us all locally and globally.

But in the bubble of “school” there is a sense that it is okay not to engage with this concretely, giving ourselves permission to continue focusing on learning objectives and outcomes tied to a past. As Michael Wesch shares in this TedTalk: From Knowledgeable to Knowledge-Able, what we teach, what we engage as “musts” as part of the learning contract are disconnected from the pressing realities surrounding us. This lack of authentic connection and scaffolding which should provide the skills and cognitive capacity to critically engage with the rapidly changing world are watered down to suit our own educational comfort zone and established routines, which have worked so well for us in the past.

Schools are keen to talk about teaching to students’ future and as George Couros share in his book ” Innovators Mindset”, its all our future and all of us need to understand and engage with it concretely, not just the kids. Education rhetoric is rich with 21 century skills terminology and on the surface we are okay to dabble with some ideas and changes, but not at the level where we truly have taken on a deep understanding as educators about what this disruption is doing to our world and how we can act as concierges of learning. We need to ensure our students are not simply consumers of the disruption but empowered creators and active participants of the disruption.

A disruption needs to occur in schools with professional development. In a video by Frank Barrett in the Harvard Business Review “To Innovate, Disrupt Your Routine” he highlights the importance of leaders ability to engage in a process where routines are disrupted to generate opportunities for creativity. As an example he uses the wonderful analogy of Miles Davis’s Kind of Blue album and the disruption he facilitated, disrupting routines which had the group enter a discomfort and then supporting them to create something new that changed the face of Jazz.

It is often the case, and I am generalizing here, that professional development in schools is focused on pedagogy framed by comfort zones that generate no disruption, learning framed with limited clear connections to the real world, lecture style delivery, and bulleted PowerPoints squeezed at the end of a long day of work. This dynamic can generate a level of disconnect, cynicism and passivity by participants, and dilutes the connection between what we teach and how we tie learning to be authentic and connected to the pressing realities surrounding us.

School leaders need to first disrupt their own professional development. They also need to be bold and challenge their own comfort zones. Through this act, we then experience first hand the process described by Frank Barrett in the Harvard Business Review, and will be willing to mentor leadership teams and faculty to go through a similar process with explicit support and care. If we are going to lead and disrupt our routines to engage in a process where we innovate and create then educators need time, space, support, empathy, and meaningful scaffolding. With this we build the capacity to disrupt their own thinking and internalize how they can take ownership with the process. We need to disrupt professional development learning outcomes, so as to be able to craft a learning narrative connected to a world driven by a new economic reality; one framed by creativity, algorithms, machine learning and innovation. Through this process we can then facilitate a culture where educators become active participants of the pressing realities surrounding us and lead and mentor learning with authentic and meaningful connections to the world of disruptions we are living in.

We cannot expect to authentically connect our students’ learning environments with the rich tapestry of the world of economic disruptions without giving ourselves permission to disrupt ourselves first and shift our professional learning discourse to unpack, synthesis, connect and craft learning outcomes that explicitly provide meaningful opportunities for us to make sense of constant disruptive change, and from there as concierges of learning, choreograph an authentic learning landscape for students. “Disrupt me!”

John@beyonddigital.org

AI Research in Japanese Enterprises — NEC

AI Research in Japanese Enterprises – Fujitsu

IBM is not the only company working on ML. Also at work are Amazon (Amazon Machine Learning), Google (Prediction API, TensorFlow), and Microsoft (Microsoft Azure Machine Learning). I will write about their offerings in the future. By the way, there are many more ML offerings from startups and small and medium-size companies as well. See Venture Scanner for more vendors for the AI/ML vendors list. Also, Dr. Adrian Bowles reported on promising companies in the AI area in his StormInsights newsletter (February/March 2015 issue).

IBM Watson Meetup (Machine Learning)

High Level Overview

The fourth speaker at the IBM Watson meetup was Dr. Shivakumar Vaithyanathan, Fellow, Watson Cognitive Services, IBM Watson. He discussed Watson declarative ML.

Dr. Shivakumar Vaithyanathan

I like the last line on his next slide, “The only constraint is the creativity of the ML Scientist.”

ML can employ many algorithms and can be described from many different viewpoints, such as which algorithms are used (conventional vs. your own). Vaithyanathan discussed it in conjunction with a distributed system (backend), such as Spark. ML algorithms work fine, as long as scalability is under control. But as more data need to be dealt with, scalability becomes a big issue. An ML algorithm that worked before may not work when data increase excessively.

Because data scientists are not trained to be experts in distributed systems, they need an easy (transparent) way to take care of higher scalability without having to know exactly how backends work.

Vaithyanathan showed this situation in the following slide. A data scientist designs and implements an effective ML for a specific need and writes a simple script in R-like or Python-like language to take care of complex backends processing. With this support, a data scientist can concentrate on his work for ML algorithm development.

Figure 1: ML and backends are interfaced with a simple interface language

In this slide, three backends are shown: Hadoop, Spark, and Open Message Passing Interface (MPI). (By the way, this article compares Hadoop and Spark.) With this interface in place, a set of necessary configurations is automatically created and pushed to the backend systems to take care of their processing properly.

During the talk, Vaithyanathan touched a little bit more on the technical side, but I’ve omitted that. I do not know whether IBM published his presentation, but I found a similar presentation by him, given in August this year, that addresses the same thing here.

More Details

If you have more time and want to dig in a little deeper, watch this talk by Fred Reiss, Research Staff Member. He now belongs to IBM Spark Technology Center in San Francisco.

Open Sourced ML

Finally, Vaithyanathan mentioned that IBM recently made its machine learning package, SystemML, open source in June this year. On November 2, it was accepted as an open source project by the Apache Incubator. More information can be found here and here.

Personal Insights

The third speaker of the meetup was Rama Akkiraju, Distinguished Engineer, Personality Insights (hereinafter, PI), IBM Watson. The talk was on one area of Watson applications.

Rama Akkiraju

According to IBM’s site, Personality Insights:

extracts and analyzes a spectrum of personality attributes to help discover actionable insights about people and entities, and in turn guides end users to highly personalized interactions. The service outputs personality characteristics that are divided into three dimensions: the Big 5, Values, and Needs. While some services are contextually specific, depending on the domain model and content, Personality Insights only requires a minimum of 3500+ words of any text.

More explanation can be found here. Its output may look like the following:

Source: IBM Watson website

Incidentally, the most recent Watson blog is Testing the Watson Personality Insights Machine Learning Model.

Tone Analysis

Akkiraju explained PI by using the example of a sales manager who is visiting a client out of town. On the plane, she needs to send an email to her sales team, whose sales outputs have been sagging recently, to tell them to boost sales. She drafts an email and runs it by Watson PI to make sure her tone is appropriate. She edits it according to its feedback.

The tone is analyzed with Tone Analyzer. Its output may look like the following.

Because she has already run her personality on PI, it knows her personality, including likes and dislikes. So when she encounters difficulties during her trip, like lost baggage, it suggests a few things to soothe her nerves.

Emotion Analysis

In relation with the tone analysis, Emotion Analysis:

uses text analytics to detect emotions from people’s digital footprints (e.g., online reviews and social media text). The service can detect emotions of “anger,” “disgust,” “sadness,” “fear,” “joy,” and also give overall measurement of happiness and intensity.

The model shown in the following slide was used for the analysis.

Further Information on Personal Insights

• Demos are here and here
• Interesting questions and answers were given in Quora, How accurate is IBM’s Watson Personality Insights application?

Summary

I think it is a good idea to apply PI to SNS and customer relations for better CRM. But if it is used to negotiate with me in areas like price (I am a consultant as well), I would feel offended. I want to think I am more sophisticated than a simple analysis by a machine.

In theory, if Watson gets enough information on me, it should be able to find an approach to please me or negotiate with me. What if I use Watson to counter my negotiator? Which one would win? After all, both sides are using Watson for negotiation. I think probably one side would win because it is not possible for both sides to obtain enough information on the opponent. I think the one with more data would win.

Finally, Akkiraju discussed Watson’s investment help. I might try this.

Cognitive Computing

Following Robert High’s Watson overview, Dr. Jeffrey Kreulen, responsible for the cognitive systems architecture for IBM Watson Group, took the stage to talk about cognitive computing. The meetup was intended to be an introduction to IBM Watson at a level that is a little more detailed than a press release. Many attendees were not experts in the subject matter, and I think the level of detail was appropriate.

Dr. Jeffrey Kreulen

He first gave the following definition of cognitive computing, asking the audience whether anyone had a different take on it. I don’t. Do you?

A cognition system:

• is able to learn through education
• supports forms of expression that are more natural for human interaction
• delivers its primary value through its expertise
• continues to evolve as it encounters new experiences and information … and does so at enormous scale

A quick search on cognitive computing reveals two descriptions:

• Wikipedia
• TechTarget

Because I am not an expert in the subject area, I cannot say too much about it. But Kreulen’s version and the TechTarget short definition appear to be saying the same thing in spirit.

TechTarget says:

• Cognitive computing is the simulation of human thought processes in a computerized model.

Another data point: according to Kreulen, the IBM Research SyNAPSE project defines a cognitive system as:

• “the quest for approximating the mind-like function, low power, small volume, and real-time performance of the human brain.”

Those who want to dig into his work more may want to read his report here. In this report, Kreulen cited several more papers and web pages, including this. There may be a more recent version of the report, but this is the best I can find.

He then covered the following components of cognitive computing:

• Elements of cognition
• Capability metrics
• Architecture
• Applications

His slide on each sub-subject was self-explanatory. Each point is explained well in this report. I intentionally did not elaborate each point, because that would have reproduced his report in essence.

Elements of Cognition (Source: Kreulen’s presentation at the meetup)

Capability Metrics (Source: Kreulen’s presentation at the meetup)

Architecture (Source: Kreulen’s presentation at the meetup)

His last slide captured what Watson is all about at a gut level, so I have attached it here.