It is solving the problem that many others are trying to solve — how to make sense of the web overrun by factory-produced, SEO-optimized diahrrea of words. Making sense of Googlesearch results has become the equivalent of playing Russian Roulette. Finding information has regressed, even as the web as progressed at neck-snapping speed.
Except D’Aloisio turns out to be a 16-year-old kid from Wimbledon, England who drops phrases like “heuristics” and “natural language processing.” After hanging out with his friends, he heads home to read papers about machine learning.
From many smart people like — Jeff Bezos, Andy Bechtolsheim, Larry Page, Andy Grove, Sergey Brin, Vinod Khosla and Bret Taylor. D’Aloisio belongs with them. Not because he has started the next hot company — who can predict what will be hot? But instead, he is a self-taught polymath, who is so adept at learning from reading, listening and observing. He is an old-fashioned technologist who was born this way.
Nick’s dad is an investment banker and his mother’s a lawyer. He grew up in Perth, Australia, where he fell in love with rugby and cricket. The clear, big night skies made stargazing a fun activity and before you know it, he was learning everything about stars and galaxies and black holes. He was not even 5. At age 7, his family moved to London. A year later, he annoyed his parents into buying him a MacBook Pro.
Why!!!
He wanted to learn how to replicate videos he saw on television.
He figured out iMovie, Final Cut Pro and some Autodeskprograms. Of
course, he couldn’t recreate the commercials he saw, but started getting
close enough. In 2007, he got an iPhone and a year later when the App
Store launched, he went out to learn the iOS SDK.A couple of months later, he created an app, SoundStumble — a geo-local music discovery app that allowed you to see what people were listening in your vicinity. Encouraged by that app, he came up with Facemood, an app that looked at your friend’s Facebook timeline and summarized what kind of mood they were in. Facemood was followed by TrimIt, an iOS app that was downloaded and used by 100,000 people. His logic for developing the app? Information overload.
“Seven months ago I got into Twitter and was getting a lot of URLs, but being on a slow phone network it would take 15 seconds to load up a page and he couldn’t view the content ahead of time,” he recalls. It was frustrating. The situation on Google is even worse. ”I don’t have enough time whilst on the go to click in and out of every article and story on the web,” he says. What he wanted was to figure out a way to skim-read its entire content before deciding to read through its entirety.
Sentiment is everything
D’Aloisio confesses that he is obsessed with sentiment analysis and each app is a step forward into his obsession with making sense of large data sets. He reads paper on natural language processing and machine learning in order to pick up the best techniques and learns from the masters. He uses iTunes University. And what he doesn’t know, he asks. He is not shy. He has emailed experts, without making them aware of his age or background.
The success of TrimIt didn’t go unnoticed. A few press mentions caught the attention of Solina Chao, a key investor with Horizons, the private equity investment vehicle of of Li Ka-Shing, the Chinese billionaire and owner of the 3 Group. Their previous investments include Skype, Facebook and Spotify. She convinced Nick and his parents that it was time for him to take the next step and turn TrimIt into a company. Of course, Horizons’ made the seed investment in what is now known as Summly, a service that allows webpages and news articles to be automatically compressed into succinct summaries.
Summly uses a text-summarizing algorithm that has been trained with sample pages from across the web and has trained itself to use optimized metrics when formulating a summary of any article or webpage. The service uses an ontological detection so that the “algorithm detects what kind of webpage has been entered into Summly e.g. a technology article, and applies our appropriate technology article summarizing metrics accordingly.” The algorithm’s preliminary evaluation showed that it outperformed other text summarizers by a factor of 40 percent, D’Aloisio claims. And it supposedly works across many different languages.
I have no idea of knowing whether his claims are accurate. Being on the road, I have not even had time to do my due diligence. The early version of the app (that is going live sometime later today) works quite well. It allows us to get summaries of articles from publications like The Guardian. We get summaries of news items via News APIs of Google and Bing. Smart tagging allows us to discover more information. It just seems to work.
“This is the right time for us to be tackling this product,” he says. Nick points out that even Google knows that something needs to be done. It has launched Google Previews, but again it is not the ideal solution. “Google’s UI hasn’t kept up with the changing nature of the web and the Internet,” says D’Aloisio and is of the opinion that the click-centric web behavior is going to soon be a thing of the past. Whether it is video, text or photos, there will be a lot more automation in how the information is surfaced to us, Nick says, arguing that it is a journey that has only just begun. “I think of artificial intelligence as applied to everyday life,” he adds.
Best is yet to come
He is learning Russian, Mandarin and a smattering of other languages. Nick, being a little impatient points out that “not being a kid.” He doesn’t want to be seen as a kid with an app. Instead, it is an app with a kid behind it.
For now he wants to focus his tiny little company on the product. He points out that most founders have a vision of the product and they can see it and when that vision isn’t brought to fruition, it causes irritation. It is one of the reasons he is still obsessive about controlling every aspect of the product. “I want to ensure it is my true vision,” he emphatically states.
From design to coding, D’Aloisio is obsessive about controlling the product — much like one of his idols, Steve Jobs.
No comments:
Post a Comment