Chatbots.org Business News

On November 1, 2011 I gave a talk to Google on natural language processing, including chatbot technology and issues in understanding simple English sentences for a story-game demo which tries to act out the sentences. Download the PDF or read full text below.

Fresh Perspectives

I am an AI research engineer. I want to create things that people use, but that stretch the boundaries of the possible. Whatever I work on, I research it and then try to come at it with a fresh perspective. A fresh perspective is something Google and I have in common, though we approach things from opposite directions. Typically my fresh perspectives involve understanding a domain and then writing a new application-specific scripting language to encapsulate insights I have gleaned. Google’s perspectives come from access to massive amounts of data and hardware.

I research everything. I even researched how to give this talk. I thought about making a lot of clever PowerPoint slides. Then I read up on using and abusing PowerPoint. I abandoned that plan.

First IBM’s Watson beat Jeopardy champions Ken Jennings and Brad Rutter. Then Apple captivated mobile consumers with the iPhone 4S, which included an enhanced version of Siri, the voice-driven assistant born at Menlo Park, CA-based SRI International. Suddenly, the idea that computers might be just as good as humans at carrying out certain types of requests seemed a lot less far-fetched.

For companies trying to win corporate and consumer adoption of their own NLP technologies, this is a long-awaited moment. And one of the firms that thinks 2012 could be the year this market really takes off is VirtuOz VirtuOz, Inc., a Paris-born company that moved its headquarters to Emeryville, CA, in 2009.

VirtuOz, Inc., the leading provider of intelligent virtual agents (IVAs) for online marketing, sales, and support, today announced that it processed a record 166 million conversations in 2011 on behalf of its customers.  This growth was fueled, in part, by expansion of VirtuOz’ client base across multiple verticals including high tech, telecommunications, retail, and consumer packaged goods.  In addition, the company noted the expansion of IVA usage from customer service, to sales and marketing via a growing number of touch points, including mobile phones and social media, such as Facebook.

Mobile devices are playing an increasing role in the lives of millions of people across the world, across all generations. As more and more online technologies deploy into the mobile realm, Virtual Agent technology is following this trend. Apple’s implementation of Siri, the voice-to-text virtual assistant feature they have added to the iPhone 4S, opened the world’s eyes to mobile virtual agent technology. Android also has added Vlingo, a similar feature that combines voice-to-text and voice recognition technology with its powerful “intent engine” to help users complete their desired action.

In a paper called Reading Digits in Natural Images
with Unsupervised Feature Learning researchers from Google and Stanford University reveal a new method of identifying house numbers in Google Street View.

It’s just another attempt to add a bit of AI to Googles ability to understand the world.

As a result however, this algorithm is obviously very useful for Captcha hackers as well, as I’ve announced many times before. Captchas are those annoying characters you have to re-type when you leave a comment on a forum, to get access to certain information or when you register for a website, The purpose of Captchas are to separate computers from real human beings, assuming that humans are better in complex character recogntion than computers. This will soon be something of the past.

The result: computers will be able to create new accounts, register everywhere, spam everywhere. It will also retrieve password, access email boxes and send emails on behalf of real humans. It’s an unstoppable development. The only way to stop this: collaborate worldwide and define a new internet which is suitable for consumer usage. Also, we should get rid of the idea that we should protect privacy on the internet allowing people to stuff around anonymously without the chance of getting caught, and that everything should be available for free. Nothing is for free.

This will be part of a new concept for the future world, which I will refer to as

The Utopia Manifesto

Soon more!

In a recent release Gartner predicted its top ten disruptive technologies for 2012. Amongst the expected and rather predictable media tablets, app stores and cloud computing, Gartner says that icons, menus and pointers will be replaced by mobile-centric interfaces such as gesture, voice and search. But voice isn’t everything. For it to be a successful method of interaction, voice needs an underlying intelligence.

CodeBaby, a company whose award-winning digital characters are shaping the growing online self-service marketplace, today announced its partnership with inbenta, a leader in Natural Language Processing and Semantic Search. Recognizing the expanding market demand for self-service solutions, CodeBaby and inbenta have joined forces to deploy intelligent virtual assistants that increase online conversation rates and customer satisfaction.

The new alliance will deliver technologies through online assistance to web visitors via the CodeBaby Intelligent Virtual Assistant (CIVA) solution which provides digital characters that engage website visitors and seamlessly guides them through the online self-service experience. By driving interaction with the inbenta semantic engine – which incorporates technologies enabling intelligent interactions though its Natural Language Processing solution – visitors dramatically increase their ability to find the information they are seeking online, thereby improving the overall customer experience. This technology reduces operations costs for inbenta customers by diminishing the dependency on customer support professionals.

Experience a foreign or distant world without actually being there: Japanese researchers leaded by professor Susumu Tachi at Tachi Laboratory have developed Telesar V Robot Avatar which delivers a remote experience straight to its operator by transmitting sight, sound and touch data. You can actually feel the shape and surface unevenness, and also the temperature of objects.

The goal of psychology is to predict and control behavior. It is the same goal we have as knowledge engineers. This may sound somewhat Orwellian, but it is not about sinister machinations. In Artificial Solutions, we want to do everything possible to help users of our Teneo virtual assistants (VAs) quickly and conveniently find the way to the information they need. Since a VA is typically designed to answer a specific set of queries, we have a clear idea of which content should be covered. For example, it’s reasonable to expect that a VA on a bank’s website will be asked questions about banking, hence it will need to be an expert on transactions such as opening accounts and transferring money, but it is not reasonable to expect it to advise you which sofa to buy with the check you write. Anna knows every item that IKEA sells, but will be puzzled if you try to borrow money from her.

Most conventional hard robots are composed of metallic structures. However, the development of soft robots can find new applications in robotics and virtual humans in augmented reality environment. An elastomeric gripper of an embodied agent equipped with an embedded pneumatic network could touch, recognize and gently pick up an egg!