What is the relationship between money and psychology?
What is the secret to happiness?
What is the meaning of life?
These are the kinds of deep questions TED Talks have long been exploring. Now, with the help of IBM’s cognitive computer Watson, the answers may be just a click away.
It maps out talks by topics and insights about the speakers, allowing users to easily browse the vast TED archives.
It also lets users ask Watson any question they can dream up. Within seconds, they receive an answer in the form of relevant clips from the talks.
For instance, when you ask about the secret to happiness, Watson serves up a series of short video segments where speakers most accurately answer the question. In one, TED speaker and Harvard psychologist Dan Gilbert says, “The secret of happiness — here it is, finally to be revealed: First, accrue wealth, power, and prestige. Then lose it.”
From there, users can click through to watch the entire talk or else ask another burning question.
The Watson-TED partnership began last winter when Dario Gil, a VP of Science and Technology for IBM Research, met TED.com editor Emily McManus backstage after giving a talk. They chatted about the possibilities of applying the powerful Watson technology to TED’s video archives.
This year, IBM developers started playing with the data. For about two months, a team of eight worked on the project under wraps, referring to the effort as Secret Squirrel.
The result is what Kai Young, IBM Watson Group program director, calls a “discovery engine.”
“It allows you to discover ideas and make connections,” Young tells Business Insider. “We can move beyond keywords to the actual ideas and insights that are part of the speakers’ content. A lot of different signals — silences, points of applause, laughter — help to understand a video, rather than just the description someone gave it.”
The program is currently in alpha testing, and IBM expects the beta version to be available by the end of the summer. Users can sign up to try it on watson.ted.com.
Developers are currently focusing on training Watson to better interpret and answer questions posed in natural language. Since it learns over time, it will continue evolving and improving.
Young sees several future applications of the technology for TED, as well as for other media providers.
For example, IBM is working on a function where users could connect their Twitter accounts, and Watson would recommend talks based on their unique personalities and interests.
Young says Watson’s technology could also someday be used to find and surface online classes, internet videos, and digital journalism.
As of now, Watson has no plans to give a TED Talk.