E-learning, Twenty Years Later

Once upon a time I was perched at the intersection of two career choices; 1) a path towards teaching and 2) a pragmatic exploitation of technology skills that we’re becoming ever more popular in the workplace. This personal nexus point occurred about 20 years in 1996. Anyone looking at my linkedin.com profile can guess which choice I made, but the story is a little more interesting than merely that of a techie who chose economic pragmatism over an academic career. Little did I know that my interest in teaching would lead me directly into the crucible of invention and innovation that has since changed the way almost everyone on the planet learns.

There are times when one realizes that he or she is the middle of something unique, something historical. I have had other experiences with which to measure that sense of recognition – for instance in 1987 when I was staying with friends in Argentina and witnessed an entire nation rise up to demonstrate support for their Democratically elected government against an attempted military coup – millions were marching in every city in the nation. The coup failed. That even was all the more significant given that Argentina had suffered under 40 years of Fascist rule until just a few years before. I tried to document it in my own way at the time but have since lost the photos I took and the notes I wrote down – what stayed with me was the sense of what true change looks like.

Fast forward a few years and I find myself studying a Masters in English Composition and Rhetoric with a concentration in TESOL (Teaching English as a Second Language). In the intervening years, I had taught English abroad as well as computer related courses and worked with web development among other things. I had both a practical and theoretical grounding in various education concepts & practices including everything from Instructional Design to Chomsky’s Transformational Grammar. All of that was interesting, but not earth-shaking per se. However, what seemed obvious at the time was that I was witnessing almost the exact moment that Education as we knew had reached a cross-roads. In 1995 and 1996, E-learning had more or less just been coined as a term and the initial preview of what was to come began to emerge. Now things like distance learning and CBTs had already been around for awhile (I had in fact helped produce video courses at community college TV station some years before), but the real game changer seemed to be the ubiquitous, global web platform & standards with the first generation of browsers like Netscape, Mosaic etc. With the web came a slew of other new technologies such as Learning Management Systems (LMSs) and collaborative meeting rooms and the beginnings of Social Networking.

In 1996, I faced a tough choice, while I felt very strongly about the importance of Education and considered it my most probable career path, I had some serious misgivings. While coming close to completing my Masters, I decided the best course would be to try to combine my academic interests with my love of technology and thus I proposed to build my Master’s thesis around emerging practice & methodology for extending E-learning into nearly every facet of traditional education. At that point, the decision was no longer in my hands – the academic committee at the graduate school I was attending rejected the proposal and in fact any thesis associated with the topic E-learning on the grounds that the topic and field were not mature enough and thus unworthy of serving as an appropriate thesis. I tried in vain to convince them otherwise; both in the context of the already significant scholarship dedicated to the topic (even in 1996), as well as in what seemed to me to be the obvious conclusion that even if the field were somewhat immature, it was all the more reason to explore and evolve it. One thing was clear – it was the future, whether the world was ready for it or not. Unsuccessful in my attempts to persuade them, I left the degree program (I got another Masters in Information Design later) and left teaching as a career to join the IT workforce. What’s more, I began looking for opportunities to help continue my original intent in helping to define what E-learning might become.

Over the next two years, I had three different projects where I worked as an IT trainer and courseware developer. This gave me both further grounding in instructional technique as well as a lot more exposure to key standards and technologies associated with E-learning. For example, my courseware projects involved creating courses on web application development – at precisely the same time that new web standards were being released by the W3C, including CSS, XML, XHTML, early forms of Javascript and the DOM. Then I got a contract at Cisco Systems and headed out to Silicon Valley. Mountain View and San Jose were quite shockingly different from Dayton, Ohio. The culture was vibrant, the Terminator was running for Governor and Silicon Valley was in the midst of the biggest boom anyone could recall – it was the age of Start-ups, IPOs and stock options - and paper millionaires (folks whose stock was valued in millions due to the massive speculative tech bubble) were literally everywhere. It was also one of the most innovative times in American history; there was an expectation that not only were things going to change for everyone, but that change would be defined in this valley.

Cisco was right in the middle of that milieu. When I arrived, the superstar CEO John Chambers had just made the announcement that E-learning was the next Internet Killer app. I even saw him walking around the giant Cisco campus a few times – I recognized him from the many magazine covers he had graced in the months beforehand. Cisco wanted to make good on Mr. Chamber’s promise and had assembled a large team of very talented people across several groups to help redefine education for the 21st Century. They tackled it from multiple perspectives, including the Cisco Academies as well as the Field E-learning Connection, a project that I became involved with. This Cisco team became a focal point for a wider group of companies, universities and institutes that began defining what next generation E-learning solutions would look like. Like all great changes though, this one was borne in the midst of some fascinating controversies and I found myself deeply involved in them.

Essentially, there were two-world views within the E-learning camp (the camp being an initial set of most self proclaimed experts and tech evangelists). While the group agreed on the shared premise that E-learning would be become pervasive across all aspects of society (e.g. fulfilling the promise of becoming the Internet’s Killer app), they disagreed what it would or should look like. The first group, who were more influential at the time, definitely had a more academic perspective and felt that Learning should adhere to fairly rigid standards and instructional design expectations. The second group (and I soon found myself within that camp), believed that E-Learning was to some extent an outgrowth of the technologies that made them possible and thus offered new opportunities to approach education in general that might prove more effective than the traditional doctrines (some of which actually date to Roman times).

This contest manifested itself in a number of ways over time, but one of the more specific examples of the battles that raged within it was the notion of Learning Objects and the delivery systems that would be used to serve them – Learning Management Systems (LMSs). The core E-learning industry was rallying around the LMS as the primary commercial Learning solution and the Learning Object standard that began driving the industry was something called SCORM (which had its origins in ISD, CBTs and the DoD). The basic idea behind a Learning Object was solid enough and seemed to conform with other emerging standards like XML – a Learning Object represented a modular, self-contained portion of a larger set of objects which be configured as needed into courses or curricula. However, this was around 2000, and the architecture behind the standard moved learning content development towards greater complexity which turned led to a higher cost per hour of content created.

Those of us on the other side of the argument certainly saw the value of SCORM based systems and content, but we had a wider view of what learning content was and how it could be delivered and managed. For about two and half years I continued participating in the debate; in the middle of which the Tech Wreck happened and I left the Silicon Valley and headed back to the Miami Valley. During these years I continued to fight for my vision of what E-learning could look like, through articles and an online community called Learning Leaders. For me it wasn’t just about business or winning a debate, the philosophical contest was very personal – to me this seemingly technical question held within it a much larger question regarding the nature of all education.

In the early 2000’s, after the tech wreck, the E-learning market nearly disappeared, just barely holding on. Then during the mid-2000’s things started to pick up again as mobile technology, portals and social networking become more prevalent. It was also around this time that learning delivery systems become more flexible and began to be adopted on an institutional scale – then quickly – almost overnight it seemed, E-learning was everywhere. Every college, every tech school and even K-12 education began hosting a myriad of Learning Technologies. Business models began changing and yet the big question was still not being addressed. The question hinted at in the dichotomy between rigid or flexible Learning Objects could be characterized this way when viewed in a larger context – should Education be flexible and learner focused or rigid and expert-driven? This is a big question – one that came up quite a lot in the recent 2016 Election although perhaps many didn’t see the connection. Essentially, anyone who was making comparisons with the Finnish Education system versus standardized assessments (a recent trend in the US), was tapping into the very same controversy. It’s a fascinating question, one that spans both personal motivation as well as the mechanisms for learning delivery (e.g. technology).

Over the past seven years in particular, the free market seems to have been moving more towards one side of this conflict than the other. Despite there still being a very rigid focus on standardized assessments in traditional environments, corporations and consumers who have been given the opportunity to choose informal or dynamic learning options versus traditional instructional design driven offerings have overwhelming moved to take advantage of informal learning. If you’re reading this on linkedin.com you can see this right now by clicking on the Learning menu item at the top of the screen or perhaps you’ve had a chance to experience the Kahn Academy or any number of similar sites. It’s been a long time coming, but we’re finally getting close to point where personal learning solutions with access to unlimited content and the ability to dynamically define one’s own courses and education paths will become ubiquitous. I think the conclusion to this story could be characterized this way – in E-learning the medium has become the message in that the medium has given us an excuse and a freedom to view Education in an entirely different way. I feel privileged to still be around to see how this field has evolved moving ever closer to its true potential and even more privileged to have been involved in helping to define what it might or ought to look like.  And that’s one of the really cool things about working in IT, because believe it or not, anyone anywhere has given the opportunity to contribute those sorts of ideas and innovation and as a result, change the world, one step at a time. I think I made the right choice…

Copyright 2017, Stephen Lahanas

Read More

A New Framework for Knowledge Management

Several weeks ago I wrote two posts on the topic of Knowledge Management (KM):

1. Whatever happened to Knowledge Management
2. Redefining Knowledge Management

This post represents the third in this series and tackles what a new framework for Knowledge Management might look like.

At the highest level, the Framework that I’m proposing for KM acknowledges several new technologies that weren’t readily available back when the notion of KM was first promoted within IT. The Framework also clearly acknowledges that KM is part of a larger ecosystem of related but separate technologies that cooperate to achieve a variety of Knowledge-related goals. Those goals could be characterized in the following manner:

• The capture of insights on multiple levels, e.g. Individual, Organizational and Community. In this context, Community represents a community of practice (probably global in nature) but could also be a market of some sort.
• The ability to both define knowledge expectations as well to discover hidden knowledge.
• Support the assimilation of source information within “levels of context” (e.g. personal, organizational and community).
• The ability to capture and reuse Knowledge Relationships & Learning Paths (I’ll describe these in more detail in a bit). This particular goal goes to the heart of one of the original value propositions behind KM in the old days – the idea that knowledge capital ought to be captured in a fashion that allows it to be reused such that if individual knowledge holders were to leave the organization it wouldn’t suffer a true “Brain Drain.”

I wanted to briefly explain what I meant by “Learning Paths” and “Knowledge Relationships.” Learning Paths are more or less Dynamic Curricula, in other words self-defining paths that traverse specific learning topics in the context of both individual and organizational learning. Let’s say you have access to 1,000 learning resources and choose to build your own learning program using say 20 of them to learn (for the sake of argument) Node.js. The chosen topics and their sequence can become a learning path which could be reused by other individuals or the organization as a whole. A Knowledge Relationship on the other hand is a little more complicated because it can be manifested in more than one way.  A Knowledge Relationship could be as simple as terms combined through metadata, or terms listed within a shared taxonomy or as complex as a SQL query or defined relationships using RDF. Knowledge Relationships is an area where the initial promise of Semantic technology has fallen a bit short of expectations but will likely continue to improve in the coming years

The most immediate realization when considering both the proposed framework and its potential goals is the fact that there isn’t now nor is there likely to be one tool that accomplishes all of what we consider to be part of a larger KM process. Maybe someday this will change, but for the foreseeable future, establishing and taking advantage of KM within an organization will require a mix of tools (many of which are likely already in place). Here is a conceptual view of the proposed framework:

There are several principles associated with this proposed KM Framework:

1. The idea that learning drives knowledge assimilation
2. The idea that analytics drive discovery and that discovery also drives knowledge assimilation
3. The idea that AI or AT can make both Learning & Discovery more effective
4. The Idea that AI or AT can empower Search capabilities in new ways and that Search drives both Analytics and knowledge discovery
5. The idea that knowledge can be layered up from the individual to the community level, thus supporting a variety of collaborative knowledge capture and assimilation capabilities  
6. The idea that underlying all these knowledge layers can reside a defined identity – this can take the shape of shared semantics, business rules and more
7. The idea that the source information (content, databases etc.) can be acted upon simultaneously from several related processes to produce meaningful insights which in turn can be captured and built upon

In many ways, I think this type of a framework is actually preferable to dependence on single class of tools in that it has a certain flexibility more or less built in. As we’ve witnessed over the past decade, there have been not one but many disruptive new technologies which can be applied to Knowledge Management including but not limited to Mobile technology, AI and Big Data. There are likely several more trends waiting on the wings that could enhance or otherwise contribute to KM in the future. This type of framework can easily accommodate any such advances.  

I will write at least one more follow-up post on this theme and in that article will explore what a real-world next generation KM process and architecture might look like within a typical enterprise and how it might be exploited in a variety of real-world scenarios.

Copyright 2017,  Stephen Lahanas

Read More

Whatever Happened to Knowledge Management?

Information Technology is a dynamic field, one often driven by buzzwords and fleeting trends. Sometimes these trends continue for decades, other times they fade somewhat quickly. One trend that experienced that fate seems to be Knowledge Management (KM). I recall first hearing about it in the early 2000’s and at that time it seemed to encompass several classes of subordinate technologies including but not limited to the following:

• Document Management
• Content Management
• Search technology (various)
• Business Intelligence (A.K.A. Decision Support or Analytics)
• Metadata Management
• Learning Management

It also tended to include more specialized tools such as knowledgebases or FAQ generators and it potentially seemed to include more integrative content-focused technologies such as Wikis. For a time KM seemed poised to also include a wide range of Semantic technologies as well. However, in the last two years or so in particular, the term Knowledge Management has seemingly dropped off of the map. I was wondering to myself why this may have happened and whether anything in particular had replaced it.

To be sure, there are a couple of related trends that have dominated the IT landscape in the past two or three years; the most significant of those being Data Science & Big Data. However, neither of these seems to fulfill the role KM was being groomed for over the previous decade. In fact, in some ways these more recent trends have become much less specific in regards to their expectations or scope (which has actually become a problem for both of them). This situation may actually help to explain what happened to Knowledge Management - perhaps the original scope was too expansive? But was it just a scope issue?

I want return again to the core or implied premise associated with Knowledge Management, that there ought to be some sort of enterprise-wide ability to help unify all of these knowledge related functions or processes and resources. The problem with this premise is related to the scope, in that there isn’t one product group of technologies associated with it, but rather a fairly large set of technologies – some of them not closely related at all other than in a philosophical sense – in other words that they could be construed as part of a larger knowledge ecosystem. So, we seem to be missing an industry impetus, but we also seem to be lacking any sort of agreed upon knowledge process or framework that would necessarily help to tie all of these diverse technologies together. This latter problem takes us deep into the heart of the larger philosophical question which KM seemed to be begging – e.g., what is the difference between information and knowledge? That isn’t an easy question to answer – and in some sense parallels my recent discussion on Artificial Intelligence versus Artificial Thought. In fact, AI could even be considered as part of KM depending on how you look at it.

So back to the tough question, what makes data or information become knowledge? Is that dependent on adding value to the data or information though specific types of processes or is it merely in the integration and analysis of such source data that the source transcends itself to become knowledge? Or is knowledge only something we can consider in a collective sense; with the sum total of all data assets being knowledge potential of some sort? These types of questions may have been raised during the years that KM was discussed actively but I think never properly answered.

You might have noticed from my initial list of related technologies that I seemed to have left out Data Management or Information Management. Either of those could potentially be considered to be part of a knowledge management framework – however the reason I left them out is that for those fields, there is a much better operational understanding of how those type of technologies work. In fact, the view from with one of these areas, Data Management overlaps quite a bit with some of what I’ve attributed to Knowledge Management (one need only look at the DMBOK to see this illustrated). Thus Data Management as a trend has continued (for decades now), primarily concerned with operational maintenance of a number of unrelated technologies without the implied necessity to integrate it all into something transcendent across the enterprise.

What if we did wish to settle the deeper question though, regarding the differing expectations between operational management of source data and knowledge exploitation? It is now 2017, are we in a position to define an integrative knowledge framework and if so what would the philosophical foundation consist of? Moreover, would coming up with this type of framework help to redeem the dimming trends of Big Data and Data Science? I think it’s worth trying to answer the question and also worth taking a shot at defining the missing knowledge framework. I will tackle both parts of this problem in two upcoming posts…

Copyright 2016, Stephen Lahanas

Read More

What We Just Learned about Grizzly Steppe

The Obama administration announced yesterday that sanctions were being placed on Russia in retaliation for the 2016 Election Hacking scandal. Shortly after that announcement, a Joint Analysis Report (JAR) was released providing a description of the nature of the Cyber attacks. It's still not clear if this report (released to scribd.com) is the complete intelligence report that the President had requested some weeks back or one perhaps one of several. What is clear however, is that the level of detail is perhaps more granular than expected, but the scope seems to be narrower than it could have been.

Architectural representation of the Election Hacks from FBI Report: JAR-16-20296

So what did we learn from the document? Here are a few highlights:

• We have a relatively straightforward diagrammatic view of how the attacks occurred (I've placed an example of this in the post image)
• We've been given a glimpse into the nature of the Russian Intelligence Service (RIS), but a limited one. Approximately two dozen names are listed as being associated with the RIS, but it's not clear if all these are indeed separate groups (and no explanation is given about any of it). There are some very Bond-like spynames in the group like CrouchingYeti, Fancy Bear and Gray Cloud but that in itself isn't very illuminating.
• We are shown some detail regarding the identity of the exploit. Unfortunately, this is not provided in a context that might be well-understood outside of the Intelligence Community or a small cadre of Cyber security experts. The exploit information is supposed to clinch the identification of the groups in question and maybe it does, however it certainly seems as though part of the story is missing.
• Fully half of the document is dedicated to describing various Cyber risk factors and mitigating actions in some detail. While this is good information, it is terribly generic and it seems a though it has been used to inflate the size of the report somewhat - perhaps at the expense of the main point for releasing it.

While I don't wish too sound too critical here, I think it might be worthwhile for the folks working on this analysis to consider creating another draft. First, I'd like to address why I think that's necessary and then I'll delve into what ought to be revised or added in the next version of the report.

The reason why we need to get this right should be obvious, but I'll state it again anyway. The report represents the foundation for both the claims that the attack occurred as well as for the sanctions that will follow. This may or may not represent a form of Cyber-warfare (both the attack and the response - I've outlined that topic in more depth here). In any case, it is a serious matter and the sanctions probably represent the most severe actions we've taken against Russia since the end of the Cold War. Thus the foundation needs to be as a strong as possible. Obviously, there are national security issues at play with this topic, however in some situations, more information can be better than less. The information missing from the current version of the report includes the following:

• Detail on the other organizations which were hit in the attack - there is an implication of a much wider attack, but no specifics.
• An explanation of the context - the goals of the attack and how the stolen information was utilized. Also, there needs to be an explanation of the process of exploit identification for those who aren't already familiar with it.
• A discussion of how the US can help safeguard election processes and systems. This is somewhat covered by the best practice portion of the report, but that seems to also be saying that all such mitigation for thwarting future attacks is entirely up to each potential target which isn't altogether satisfying. We should be having stronger a dialog on how critical processes can be protected by the groups we thought we there to perform that task. For example, who if anyone, will take the lead on auditing voting systems in every state?

The current Grizzly Steppe report seems to have give us the bare minimum. We need more than that if we wish to learn from this experience and keep it from happening again. Let's give it another try...

copyright 2016, Stephen Lahanas

Read More

A Framework for Evolutionary Artificial Thought

This past week, I was making the long commute between Dayton and Columbus, Ohio and trying to amuse myself the best I could – in this case by listening to a college course on the fundamentals of Particle Physics. One might not think there is an obvious connection between Particle Physics and Artificial Intelligence, but it turns out there is at least one. The connection, in my mind at least, was the framework used in Physics to help organize the field of subatomic particles – it’s known as the Standard Model of Particle Physics.

In the Physics world in the early 20th century, as in the field of Artificial Intelligence now, there was a tremendous amount of information which was obtained from various sources and very little ability to place it all into a unified context. We discovered particles, defined new mathematics and reinvented philosophy through Relativism and Uncertainty, yet without a guiding framework it all must have seemed terribly chaotic and random for those working in the field. Out of that chaos emerged a framework though; half invented, half discovered – one that helped to focus technology, philosophy and application of the science in question. I thought to myself, this is exactly what we’re missing with Artificial Intelligence. But as I posited in my last post, the framework that’s needed might also require a philosophical adjustment – one that illustrates Intelligence in the context of a combination of evolutionary capabilities which might best be described as Artificial Thought. In that article, I tried to make the philosophical case for something like that might make sense, in this post, I’m going to get a bit more specific and examine the more pragmatic aspects of a Framework for Artificial Thought.  

Let’s start at the beginning – what does this or any such “Framework” buy us? In the case of The Standard Model of Particle Physics, the framework provided the following benefits:

1. The ability to place a number of potentially divergent concepts and discoveries within a unified and coherent context
2. The ability to explain the nature of something in a manner consistent with empirical data
3. The ability to support a variety of predictions, like for example the discovery of specific types of new particles, the most recent and famous of those being the Higgs Boson

For Artificial Thought, the benefits might be a little different, but perhaps not as much as one would think. The high-level value proposition associated with a framework for Artificial Thought might include the following benefits:

1. The ability to align a diverse set of AI theories, techniques and technologies within a coherent, unified context
2. The ability to define a clear evolutionary path within that context whereby component capabilities can be combined to achieve ever greater orders of Thought and eventually Intelligence
3. The ability to better chart and predict success in achieving Thought or AI milestones

I suppose the biggest difference between Particle Physics and Artificial Thought is that we’re bypassing the need to discover the functions of natural intelligence on the biological level. In other words, there are no CERN-like labs available to discover Thought in progress the way we discover subatomic particles. This could make our efforts harder to achieve, but perhaps only if our goal was to recreate natural intelligence as opposed to generating capabilities which are logically similar if not actually organic in nature. This brings us back to the central premise in the previous article, that recreating the most complex and comprehensive capability is a hell of a tough goal and we should instead worry mostly about intermediate steps rather than the end game (and I might add without getting lost in the growing tangle of 100’s of immediate, lower level details or approaches and opportunities)

Let’s take a look at what a Framework for Artificial Thought might look like…

The framework resembles an IT architecture because in a sense that’s what it is. Imagine a situation in a few years where we have a galaxy of AI-related capabilities, how might they work together and to what end? This view gives us a hint as to what that might look like. It does more than that though, it also shows how we move from lower-level Thought to higher-level Thought and it also begins to illustrate the potential for orders of Artificial Intelligence through integration of Thought capabilities (both within and across Tiers). The Tiers themselves mimic to some extent the natural intelligence we’ve referred to (both human and otherwise) by illustrating how basic capabilities might evolve into something more.

Tier 1 is Awareness and I think it’s safe to say this is the area where traditional AI has made the most progress thusfar and that only stands to reason. As we discussed before, Awareness in this context is nothing at all like Self Awareness. We can imagine a rover crawling along the rocky, barren landscape of Mars, avoiding obstacles by becoming aware of them through sensory apparatus and that fits this tier just fine. Is the rover Intelligent? Not really, yet some of the rovers we’ve built or are building can potentially operate on their own without explicit direction or intervention from human operators. This is a good starting place…

Tier 2 might witness our rover becoming more sophisticated, perhaps learning from its environment and building upon its experiences yet still basically reacting to environment. As you might have noticed from the diagram, it’s clear that Artificial Thought will occur both individually and collectively – something perhaps not fully anticipated by the founders of AI back in the 50’s and 60’s. This has critical implications to the applications of Artificial Thought, for example in the case of the rover, the implication might be that it distributes some of its higher function elsewhere. If maintaining higher Thought from Earth presents difficulties due to the 8 minute travel time for instructions, then perhaps there might be cognitive capability in another part of the lander or on an orbiting satellite. The key idea here though is that the rover itself doesn’t need all of the cognitive capability itself, which would become even more important if for some reason the rover was instead some type of Martian UAV and needed to operate for long periods of time with minimal fuel. The bottom line, though, is that Collective Thought is still Thought, regardless of how it might be distributed or otherwise combined. This is one area where our current view of individualistic human-mimicked Thought really diverges from where we’re headed.

The question begins to arise in relation to Tier 2 capability as to whether if we combined all of these functions and integrated them somehow, would we in fact achieve a level of Intelligence? For the sake of argument, let’s say yes, that it would. If we combine all capabilities that fit with Tier 1 and Tier 2 Thought, we might say we have achieved a “level 0” order of intelligence. I don’t think Tier 1 by itself would justify that assignment, yet when we look at what Tier 1 represents it does seem to mimic much of what might be required for a lower order of intelligent life to survive.

With Tier 3, things get more interesting. This is where Watson and some of the other more ambitious AI projects have been focused with limited success, but still some progress has been made. The distinction between reactive and proactive, between simple and complex is a big leap and one that not all Artificial Thought has to make. It’s important to keep in mind here also that the Tiers are not actually separate from one another, rather the higher tiers build from the lower. This model is evolutionary on several levels, both in terms of building capabilities up but also in our ability to mimic the progression of Thought metaphorically from its beginnings to somewhere at least close to how we view it. Beyond those considerations, it also represents a real-time integration architecture as well – with lower information feeding higher level capabilities. If enough integration (across Tier 3 capabilities) occurs then we might say that we’ve reached a “level 1” intelligence.

Of course, I haven’t defined what level 0 or 1 orders of intelligence represent, but the taxonomy might look something like this:

• Level 0 – An order of intelligence mimicking primitive life
• Level 1 – An order of intelligence mimicking intermediate forms of life, but not humans
• Level 2 – An order of intelligence that truly mimics human intelligence
• Level 3 – An order of intelligence beyond human intelligence

Each of these orders or levels of Intelligence involve a multitude of complex Thinking behaviors (I’ve abstracted the view to a great extent for this dialog). The framework I’ve outlined above isn’t focused on level 2 or 3 Intelligence, we can leave that for science fiction for now. But, the next ten years could see some remarkable breakthroughs on the lower levels.  “It thinks, therefore it works,” might be a good corollary to Descartes’ original premise.

Copyright 2016, Stephen Lahanas

Read More

How Artificial Thought Can Save AI

Over the past several decades, I’ve seen an endless stream of predictions and articles in regards to Artificial Intelligence and it occurred to me recently that we may have missed an important point relating to this topic. One reason the expectations and the reality of AI have diverged so greatly may be due entirely to our obsession with the notion that in creating it we ought to be somehow be mimicking ourselves through some sort of human intelligence without perhaps truly understanding what that represents. This seems to be simultaneously our greatest goal and our worst nightmare relating to AI. But then I got to wondering, what’s the difference between Artificial Intelligence and Artificial Thought and if we viewed the question from the latter perspective are we in fact actually making some real progress?

Before we can dive into that question it is worthwhile to try to define what we mean by Intelligence and Thought. Here’s a good definition of Intelligence (signed by 52 scientists in the field):

A very general mental capability that, among other things, involves the ability to reason, plan, solve problems, think abstractly, comprehend complex ideas, learn quickly and learn from experience. It is not merely book learning, a narrow academic skill, or test-taking smarts. Rather, it reflects a broader and deeper capability for comprehending our surroundings—"catching on," "making sense" of things, or "figuring out" what to do.

I like this definition as opposed to some others I’ve seen because I think it carries within it a larger scope that we tend to include with the notion of what Intelligence is or what an intelligent being / entity has to be able to accomplish. This type of definition serves us for example when considering what extraterrestrial intelligence might be. Thought, at first glance, might be considered a subset of Intelligence – it is the act of demonstrating one’s Intelligence per se but also the product of that demonstration. The definitions of Thought are somewhat less concise and often seem a bit recursive though – take this one for example: “a single act or product of thinking; idea or notion.” Thought can be both a verb and a noun. Perhaps the reason that it is so difficult to nail down the definition for Thought is because we do tend to view it as a subset of the superset Intelligence and describing a component of that process or capability without fully explaining or understanding it is challenging. One definition that I think fits our topic a little better comes from the Merriam Webster dictionary; “reasoning power or a developed intention or plan.” This definition of Thought doesn’t try to explain it as much as it tends to highlight how or what it represents. Interestingly, this more compact definition also closely mirrors some of the first definitions for Artificial Intelligence, but we’ll return to that in a minute.

So, at the highest level, “Intelligence” might be considered a higher reasoning power, one that also tends to imply self-awareness and continuity of thought and a sort of assimilation of knowledge into a Self as time passes. Intelligence may be more than that as well in that there is an integrative aspect to it that often isn’t included the definitions – sometimes we view that integrative aspect as Self but not always. If we return to the Ontology of the subject, we might be able to state that you can’t have Intelligence without the thought process or individual thoughts but perhaps you can have thoughts without Intelligence per se. In other words, Intelligence at least on the face of it, seems to represent a higher order than Thought, although as we know from the real world that there are also various orders or levels of Intelligence too.

This is more of a philosophical question then a technical one, but let me follow it a bit further. Let’s say that there can be Thought without Intelligence and the key difference between the two in the context of Human Intelligence (which seems to be what many have tried to emulate) may be self-awareness and the ability to integrate reality within a unique perspective and context. At its most fundamental, Thought can be disconnected from other Thought as well as from any experience or capability resembling self-awareness and / or complex integrative interpretation. By this definition, there are likely a wide variety of animals that have mental activity that might be described as Thought, but certainly not Intelligence in the sense we tend to attribute to humans. There can be lower levels of Intelligence, but the difference there between Thought has less to with self and more to do with Integrative interpretation. If we view any organism as a complex system or system of systems, then some level of integrative coordination is always occurring. This coordination is often automatic, even for humans, but sometimes it is deliberate. Deliberate integration may be classified as Thought, but it doesn’t necessarily require Self-Awareness.

All of this begs the question, what is Self-Awareness and why is it so important in the distinction between Thought and Intelligence?

Self-Awareness implies an understanding or expectation of identity. There can, of course, be Awareness without any indication that a Self exists. Take away Self from the equation of Intelligence and what you have left are a lot of the same capabilities; such as Learning (to a point), Memory (from an objective rather than a subjective context), and Thought which can accomplish many of the same goals that intelligent Thought can, but not all of them. Take Self away and you can also potentially discard the requirement for complex integration (the level and type of integration can become much more selective). We might refer to this Selfless state as “Targeted Thought” and this is a little more interesting in that it represents areas within which Thought can be developed, specialized and reinforced without higher-level expectations. The difference between that and Thought in the traditional sense (as a subset of Intelligence) is that there are likely to be clear boundaries that constrain the operation of Targeted Thought. A Targeted Thought “Boundary” for example might operate solely within the context of airline routes and all of the processes directly associated with flight routing. Within that boundary, reasoning power based on planning and guided toward specific intentions could take place to help solve questions of efficiency or profit. This type of Thought can still be considered “novel” thought as long as it isn’t fully determined in advance, but it is unlikely that something artificially constrained to a single purpose ought to be considered intelligent in the way we tend to view Intelligence.

So, Artificial Thought in the abstract sense might be considered as the ability to derive novel outputs from similar inputs based on real-world situations and unique (yet mostly static) rulesets and perhaps various Targeted boundaries. This definition provides a framework in which Artificial cognition or Thought might obtain near-term success – a much narrower view to be sure and one already potentially aligned with nearly every practical AI effort yet undertaken. This definition is perhaps not too far off from what the initial definitions for Artificial Intelligence were, yet the types of predictions that we can make about what Artificial Thought can or can’t do will likely become much better defined within this narrower confine.

Let’s go from the abstract to the real world, Nature. In Nature, lower level organisms likely have many potential applications for Thought but don’t build that type of Thought around language or complex symbols but rather through some sort of connection to various sensory capability and stimuli. It wouldn’t be expected for a squirrel to memorize a complex path between several dozen trees bearing acorns, yet in order for the squirrel to succeed and survive through the Winter, he must have a well-defined ruleset informed by recent and current experience that facilitates such navigation. The squirrel’s journey is a problem-solving exercise, one that he may or may not be able to learn from or remember but one that he has to be able to repeat successfully under dynamic conditions. The squirrel might be viewed as a system which demonstrates a limited level of awareness, and employs novel thought in an integrative manner so it might be considered a lower-level intelligence (when all of those capabilities are combined).

My point with the squirrel analogy is this; if we were to attempt to create & program an Intelligent squirrel (using the more expansive of scope of the definition for Intelligence) to think through all of this we might be missing the point or be conducting a certain level of overkill. We could instead take any portion of the capability the squirrel possesses and use that to solve various types of problems or perform tasks. In other words, in deconstructing even a lower order of Intelligence, we can extract some Artificial Thinking capability that might prove useful – capability that likely far outstrips what we’re currently able to do. 

And even if we were to view all those capabilities combined in the case of the Squirrel’s survival, he simply doesn’t require a general purpose Intelligence the way we’ve been defining it in AI, but he does require a certain level or type of Thought (novel action resulting from dynamic inputs). This is not to downplay the complexity of the squirrel’s mission, which is in fact fairly daunting. Any successful squirrel must regularly evade predators, find food, find shelter and navigate its environment in a dynamic 24/7 environment. The navigation problem alone is challenging enough, and we use such problems all the time to test the efficacy of AI programs. However, while the squirrel needs to be able to jump from one branch to another without falling, it doesn’t require a complex understanding of Physics to do that successfully. A squirrel utilizes sensory information to determine speed, distance and other factors and makes a decision as to whether he should or shouldn’t jump. If we can achieve any sort of novel decision making, even in areas much less complex than the squirrel experiences, we will have made serious progress.

Let’s return to the concept of Thought again for a moment. What we’ve just described as Artificial Thought in essence represents a combination of Data Fusion (sensory data) and logical problem solving. The rulesets can be hardwired directly into the thinking machine, without it ever having to learn or improve upon them, although some limited form of learning may be a possibility. The requirement for natural or machine learning is not an absolute necessity for Artificial Thought whereas it might be for AI. Nature for example, endows its various creatures with a minimal number of rulesets – just what’s needed and not much more and only very basic learning abilities. This primitive level of Thought and corresponding lack of self-awareness is perfectly acceptable for a multitude of tasks (and keep in mind that a lack of self-awareness does not imply lack of awareness). This scenario, or even pieces of it, still more or less surpasses what Artificial Intelligence can achieve today, yet it represents a much more realistic target if we reconsider how we might go about achieving it. If we were to combine a number of diverse Thought processes we might be said to be building a lower order of Intelligence, but not yet perhaps the general purpose, human inspired intelligence most associated with AI. And that is perfectly ok as it still represents real progress.

There’s another consideration and area of confusion as well when it comes to the struggle to create Artificial Intelligence; it has to do with the obsession of creating architectures inspired by what we believe to be the structures underlying human intelligence (e.g. the human brain). Neural networks as a metaphor is perhaps the best example of this but not the only one. Sometimes, I think this is like designing a mission to Mars based on the understanding of how a bottle-rocket works; while there are bound to be some similarities – our understanding of human physiology in regards to Intelligence is still relatively primitive. More importantly perhaps, is the realization that there are a massive set of applications available for machines that think but aren’t necessarily intelligent. This means that even if we did understand how to recreate human or even a general intelligence, the overhead for doing so might not really be necessary or at least not yet. There’s a lot we could do using mere Thought in a more selective sense.

And why is selective or Targeted Thought of any value? Well, if we look at IT for example and the amount of effort directed towards explicitly defining behaviors in code the value should become readily apparent – Thought – even targeted and selective Thought – could make the operation of any type of machine or system infinitely more efficient. And just because the machines in question cannot appreciate the Thoughts they’re having, doesn’t mean that can’t build new behaviors or learn from an initial set of foundational rules (either individually or collectively). We don’t have to worry about recreating nature per se, we simply have to keep in mind the pragmatic motivations behind the value proposition that nature has illustrated so convincingly to us. If we do that, and work towards simpler goals in an evolutionary fashion, we can grow Artificial Thought into a powerful part of most industries. This approach also builds upon areas where success has already occurred and has the potential to accelerate those successes but it also tempers expectations in regards to what can or even what should be done. Artificial Thought (as opposed to Artificial Intelligence) can become much more focused and specialized in terms of its architectural objectives.

It’s time to come full-circle and consider again why there is such confusion and disappointment in the field of AI. It starts with the definition for Artificial Intelligence:

Artificial intelligence (AI) is the intelligence exhibited by machines. In computer science, an ideal "intelligent" machine is a flexible rational agent that perceives its environment and takes actions that maximize its chance of success at an arbitrary goal. Colloquially, the term "artificial intelligence" is applied when a machine mimics "cognitive" functions that humans associate with other human minds, such as "learning" and "problem solving." Here’s another definition:

A branch of computer science dealing with the simulation of intelligent behavior in computers; the capability of a machine to imitate intelligent human behavior.

When I refer to Thought versus Intelligence, I’m not implying that Thought can or should resemble human thought, rather what I’m positing is that Thought is any cognitive ‘processing’ that isn’t explicitly programmed up front (e.g. it is semi-random in nature based upon inputs fed into it). We’re not talking about simulating human intelligence or behavior and this type of thought is limited in terms of the Intuitive or contextual capability – it is certainly not creative Thought. The distinctions I’m making are important. In our industry (IT), we’re perhaps too enthusiastic in pronouncing this or the other technology as being “Intelligent” in some regard. The reality is that none of them truly are, which is also why after pursuing Artificial Intelligence for more than 50 years, few are willing to say anyone has actually achieved it. But that’s not to say we’ve achieved nothing – we have in fact built a wide variety of foundational technologies which are coming close to filling effective roles as Cognitive Aids. These technologies don’t simulate human thought, but rather expand or supplement it with Artificial Thought which we can choose to apply intelligently or otherwise. The type of thinking that we should be focused on is discrete, focused and targeted.

I supposed this dialog risks the possibility of replacing one vague and hard to realize concept with another, but there does need to be a way to classify intermediary cognitive capability that goes beyond standard computing but falls short of human cognition. We don’t need to speculate too much on the quality of thought in various animals that clearly have the ability to think on some level and we might extend the same courtesy to machines or systems. We don’t have to consider either as Intelligent to appreciate some value in what they do – and many successful organisms obviously don’t think at all – but the ones that do can become role models so to speak for the near-term goals associated with artificial cognition. Understanding or recreating biological neural processes aren’t necessary here either – the models we’re aiming for are pragmatic and logical with animals simply providing a useful analogy if nothing more (which means we’d use them as models in a way somewhat different than we might for Robotics). 

Descartes once said, “I think, Therefore I am.” Someday, I’m sure there will be a machine that becomes Intelligent in this context, by becoming self-aware. It’s time to recognize that the path towards machine intelligence ought to follow a more rigorous evolution of less lofty goals. In my next article in this series, I’m going to provide a framework for classifying types of Artificial Thought and discuss how that can be allied with architectural objectives as well as current or near-term technologies and applications.

Copyright 2016, Stephen Lahanas

Read More