2 Comments

Summary:

We’re entering the age of the smart personal assistant, as computers increasingly listen and understand what we’re saying and fulfill our requests in real time. Siri and Watson have gotten a lot of attention but we’ll see this type of service in a variety of areas.

header

When I first heard of Zemanta, I thought of it as a great tool for bloggers, helping recommend links, content and images. And it does that quite well, helping some 80,000 active users.

But after talking with Zemanta’s CTO and co-founder Andraz Tori, our conversation turned to the bigger picture of what New York-based Zemanta is doing. And it’s really in a similar vein as Apple’s Siri, IBM’s Watson and other services. We’re now entering the age of the smart personal assistant, as computers increasingly listen and understand what we’re saying and fulfill our requests and questions in real time.

Siri and Watson have gotten a lot of attention but the interesting part is what’s happening behind the scenes. Computers are increasingly using natural language processing, semantic analysis, and machine learning, combined with loads of data to “understand” humans, conjure up answers and recommendations or direct people to services and information.

Tori said we are going to start seeing this type of service in a variety of areas, providing a personal level of computing that hasn’t been possible before. It’s not just going to be in smartphone apps like Siri, but woven into all kinds of services.

“We used to treat computers like a tool, something that helps you do a task. Now computers are moving into the area of coaches, assistants or collaborators,” Tori said. “Siri is just the beginning. We will see a personal career advisor with you being able to explain where you want your career to move and it will say you need to meet these people. Or it will tell what you should eat today and how to plan your health.”

Tori said technologies like the one being used in Zemanta show how personal assistants can serve a wide variety of purposes. Zemanta works as an extension for browsers or a plug-in for blogging platforms such as WordPress (see disclosure), MovableType, Live Writer and others.

The technology reads while a writer taps out their post and suggests in real time relevant links, images, text and keywords to help complete and enhance a story. The power of Zemanta is that it’s understanding context from what a user is writing and then consults huge data sets pulled from sources such as Wikipedia, IMDb, Amazon and background knowledge that the computer has learned about the world. It starts spitting back suggestions and also learns by what recommendations a user applies and doesn’t.

Tori said it’s the ability to handle big data that is one of the enablers behind the rise of personal assistants. These personal assistant services are able to handle tons of information and are learning to form answers by consulting multiple data sets. Zemanta, for example, knows about 7 million concepts and topics that it’s learned using semantic, linguistic and statistical data and how people are actually using the terms.

“The hard part isn’t using just one signal but to use all the data all together to put together a suggestion. It’s a mix. You have to have real time processing, have a background database and understand signals from devices and the Internet. If you join that together you have something new and valuable,” Tori said.

We’re still just scratching the surface of what’s possible. Systems like Siri are still limited to certain tasks and domains. Watson is slowly moving beyond winning Jeopardy challenges to looking at applications in health care and retail. Some of the breakthroughs will be in making the interfaces more robust, improving speech recognition for example or taking in more data from mobile devices. We’ll also see this applied to a broader range of domains, such as potentially detecting lies from politicians.

But overall, we’re transitioning to a world where computers will be able to move beyond understanding words and analyzing surface interests and past habits. They’ll increasingly analyze our meaning and intent and will provide timely responses, augmenting our lives much like a real personal assistant.

Disclosure: Automattic, maker of WordPress.com, is backed by True Ventures, a venture capital firm that is an investor in the parent company of this blog, GigaOm. Om Malik, founder of GigaOm, is also a venture partner at True.

  1. We are using Zemanta to parse short updates from our users to extract context and key elements. Zemanta is not perfect yet for short updates but is the best solution out there. They rock when it comes to finding context and key elements in blog. We absolutely love their technology and building on it. Zemanta has a great team and we are sure they will nail the solution for tweets and short updates too! Highly recommended for bloggers and language parsing applications.

    Share
  2. NLP enabled agents will start being added to apps and services just as social was added these past five years. For example, check out this office productivity tool http://www.dynamicalsoftware.com/ccm/beta where an assistant listens in on a chat room discussion then publishes the summary in a CMS.

    Share

Comments have been disabled for this post