What Does Cognitive Computing Mean for Virtual Agents?

Watson2Forbes recently published an article on IBM Watson and the intent to use Watson’s cognitive computing abilities to provide customer help desk services. In a previous post, I wrote about the technology behind DeepQA, which is the question answering framework used to power IBM Watson. The DeepQA technology seems like it could be a game changer for the virtual agent landscape.

How does cognitive computing push the envelope? DeepQA / IBM Watson is not constrained by the limited amounts of information that most current virtual agents have access to. This is an evolutionary process. Chatbots have been far outstripped by intelligent virtual agent technologies over the past decade. Chatbots are simple pattern matching devices that can only respond to a question that has already been programmed into their database. Virtual agents don’t have to be primed with all the questions in advance. They have the benefit of speech and/or language recognition, natural language processing, and search. They can understand a person’s intent and search a limited database or set of website content to find an answer, primarily based on keyword matching.

Cognitive computers like DeepQA go even further. Such question answering systems can be primed with almost unimaginable amounts of information. IBM Watson can be given access to every bit of internal product and company documentation, as well as online review sites, analysts articles, and on and on.  This is far more information than a traditional virtual agent, or even a human, could consume and process. Relying on its massively parallel probabilistic evidence-based architecture, IBM Watson can very quickly find possible answers to almost any question, determine which answers are most likely to be correct, and offer a response.

What do these newly emerging question answering technologies mean for the future of customer service virtual agents, personal digital assistants, and web self service as a whole? It remains to be seen how products such as an Ask Watson will perform in the real world. It’s also not clear if Ask Watson will be cost competitive when compared with more traditional solutions.

How the virtual agent landscape develops will depend not only on emerging technologies, but on what consumers expect from the systems they interact with. We’ll be keeping an eye on developments.

Ray Kurzweil’s Ambition and Musings on The Future of Virtual Agent Technology

Digital BrainSingularity Hub did an interview with Ray Kurzweil back in January, during which Kurzweil talked about his vision for an artificial intelligence that will act as a trusted personal assistant to humans. Kurzweil had only just started his stint at Google when the interview took place. He briefly shared his vision of constructing an artificially intelligent software system that mimics the hierarchical architecture of the human brain. It remains to be seen how successful Kurzweil and the team at Google will be in their endeavor. Whatever the outcome, the race to produce smarter and smarter digital entities is definitely underway. As Gary Marcus points out in his review of Kurzweil’s book on building a brain, there are many different machine learning techniques and cognitive systems that are being researched today in the public and private sector. Whether Kurzweil’s hierarchical approach pans out or not is really irrelevant. Advanced AI that can understand human intent and provide answers to human questions will happen. The progress made in this field is bound to influence commercially available virtual agent technologies, both in the mobile personal assistant and in the enterprise and customer support virtual agent domains.

Older conversational agent technologies will most likely be superseded by new ones. The work that Ray Kurzweil, the DeepQA team, and many other artificial intelligence researchers are engaged in today is producing techniques that far outpace the rudimentary pattern matching technology deployed in most simple chatbots. In the hands of dedicated and savvy bot masters, chatbot scripting languages such as AIML can be used to create impressive question answering agents. But unless it is combined with natural language processing, search, and machine learning algorithms, AIML by itself can’t produce a truly effective virtual agent.  It’s sporty to make any predictions when it comes to the future of artificial intelligence, but one pretty safe prediction would seem to be this: the intelligent virtual agent that one day passes the Turing Test won’t have been created using basic AIML pattern matching technology.

For commercially viable virtual agents in the field of customer support, incorporating strong search capabilities would seem to be a must. Search can be combined with pattern matching against a broad database of known frequently asked questions to provide web or mobile users with basic self serve information. Text and/or speech recognition and natural language processing would also seem to be non-negotiable skills for a virtual customer service agent. User profiling and targeted recommendations are capabilities that advanced virtual agents should also have in their toolkit. We could go even farther and list attributes such as a sense of humor, the ability to detect human emotion, and empathy. All of these would be desirable qualities in a customer-facing virtual agent.

Perhaps as Kurzweil / Google and others work towards recreating the human brain in digital form, advancements in cognitive computing, speech recognition, natural language processing, and other interrelated fields will be the outcome. It will be hugely interesting to see how software vendors in the virtual agent and personal digital assistant space capitalize on these breakthroughs to improve and reshape their commercial offerings.

What Makes IBM’s Watson Tick? And Is Watson the Future of Virtual Agents?

WatsonI ran across an incredibly detailed and interesting article on the technology behind IBM’s Watson, the cognitive computer that outplayed several human Jeopardy champions back in early 2011. (By the way, do you know who Watson is named after? I thought I knew too, but I was wrong. I’ll give you the answer a little later).

I highly recommend the article for anyone interested in virtual agent technology, cognitive computing, or artificial intelligence. I won’t do it justice, but here’s my cliff notes version of what makes Watson work. Watson is an outgrowth of the IBM DeepQA project, where QA stands for Question Answering. IBM and other research groups founded an open source initiative called Open Advancement of Question Answering (OAQA), and OAQA continues to work on building and improving upon a common methodology and architecture for automatic question answering systems.

The article describes DeepQA as utilizing a “massively parallel probabilistic evidence-based” architecture. To break that down into something less than geek speak, DeepQA has the ability to quickly categorize and understand a question and access lots of varied data sources to pick out many instances of information that might answer the question. It can then sift through all the answer candidates to rank them in terms of which answers have the highest probability of being correct. Only when it concludes that the possible answer set has a high enough probably of being right, will it attempt to answer the question.

How DeepQA navigates its way through each of these steps involves many complex processes. It first has to analyze and understand what’s being asked. It uses techniques such as identifying the lexical answer type, which means recognizing an important word in the question or clue that points to the type of answer. For example, in the jeopardy clue “The only President not to live in the White House,” the word “President” identifies the type of answer. (The correct response, by the way, is “Who is George Washington.”) The DeepQA system analyzes the clue in detail using knowledge of specific patterns, key words, lexical answer types, semantic relationships, and many other factors. Once the system has grasped the meaning of the clue, it uses shallow and deep searching techniques to comb through copious stores of content to identify possible answers. Interestingly, some of the most complex behaviors that DeepQA performs occur after it has gathered up all the potential solutions. It gathers supporting evidence and uses sophisticated ranking algorithms to determine how likely each answer is to be the correct one. See the full article for a much more in depth view of the core technologies that power DeepQA.

These abilities will certainly influence the future of virtual agent technologies. Customer-facing virtual agent support apps are, after all, basically question answering systems. A customer asks a question and the virtual agent needs to be able to quickly provide the best possible answer. The beauty of the DeepQA and OAQA framework is that it has a much better chance of understanding the customer’s question than a more limited and simplistic pattern matching system. The OAQA architecture should also easily outperform virtual agents that search website using key words in order to respond to user inquiries.

Will the technology of DeepQA be available for customer support and enterprise virtual agents in the near future? We’ll be watching the marketplace to find out. Also see our recent post on IBM’s Watson Goes to Med School. Oh yes… and the origin of Watson’s name has nothing to do with Sherlock Holmes. This cognitive computer with a knack for solving tough puzzlers was named after the founder of IBM, Thomas J. Watson.