Get our free guide and buy smarter. Shop smart. Eco-friendly alternatives to plastic toothbrushes Can they really keep both your teeth and the planet clean?
The best tasting vegan chocolate Which dairy-free delight was judged the best tasting by our expert judges? Why fabric softener is a waste of money Fabric softeners are unnecessary and expensive, and some don't do much more than plain water.
Summer fun and safety. Tesalate full sized 'sand-free' beach towel review Can Tesalate's towel really keep you sand-free and get you dry at the same time? How to buy the best pedestal and tower fans Not just spin: some fans are better than others. Nine tips for a sparkling clean barbecue Tried-and-true barbecue cleaning tips for backyard grillmasters, including the TikTok hack you should ignore. Best beach shelters to keep you sun-safe this summer The pros and cons of BYO shade, plus beach etiquette tips.
What to look for in a rashie How to choose the right sun-smart rash vests to protect your skin from the harsh summer rays. Making change together How you've helped us make things fairer for Australians.
Learn more. OCTober October JULY She looks at the floor as she walks and avoids eye contact with customers. Geitgey chimes in, suggesting she wasn't popular in high school and is shaking off her past by working in a cool place. Kuhlke cuts him off. So what? She's cool enough to know she should pick Audrey Hepburn, but not cool enough to pick the right movie, 'Roman Holiday.
The waitress is obviously attractive, maybe the prettiest woman in the restaurant. And she does seem bothered by - or at least indifferent to - her surroundings.
But that hardly makes her prone to ill-informed choices. Kuhlke reconsiders. Maybe she would like 'Roman Holiday. She's disaffected, he insists, throwing his arms in the air. Just back from a visit to his childhood home in Kiev, Max Levchin is in his South of Market office holding up a propaganda handbill that he picked up in Moscow. He wants to mount it on a wall at his new startup, Slide. The poster features a railroad worker staring out from a caboose and a headline that Levchin translates aloud.
Levchin's team doesn't really need the reminder. The year-old CEO drinks eight espressos a day and possesses a work ethic that's famous around Silicon Valley. He often arrives in the office by six and is still there 12 or even 18 hours later. Levchin cross-referenced traffic patterns, auction histories, user interactions, geography and a thousand other factors to root out fraud.
I'd have printouts and printouts graphing the relationship between any two variables," he says. These days, Levchin has the typical angst of a second-time entrepreneur. He's obsessed with proving that his first success wasn't a fluke, and he wants to exploit the biggest opportunity he can find.
He's using what he learned at PayPal, not to root out fraud but to create the best recommender system he can imagine, one that will cover the entire Web, pulling content of all kinds - music, movies, gadgets, blogs, news stories, cars, one-night stands, you name it - filtering it according to individual preference and delivering it to the desktop.
Instead of quantifying the odds of your stealing money, he's building a "machine that knows more about you than you know about yourself. If Slide is at all familiar, it's as a knockoff of Flickr, the photo-sharing site. Users upload photos, which are displayed on a running ticker or Slide Show, and subscribe to one another's feeds.
But photos are just a way to get Slide users communicating, establishing relationships, Levchin explains. The site is beginning to introduce new content into Slide Shows. It culls news feeds from around the Web and gathers real-time information from, say, eBay auctions or Match. It drops all of this information onto user desktops and then watches to see how they react.
Suppose, for example, there's a user named YankeeDave who sees a Treo scroll by in his Slide Show. He gives it a thumbs-up and forwards it to his buddy" we'll call him Smooth-P. Slide learns from this that both YankeeDave and Smooth-P have an interest in a smartphone and begins delivering competing prices.
If YankeeDave buys the item, Slide displays headlines on Treo tips or photos of a leather case. If Smooth-P gives a thumbs-down, Slide gains another valuable piece of data. Maybe Smooth-P is a BlackBerry guy. Slide has also established a relationship between YankeeDave and Smooth-P and can begin comparing their ratings, traffic patterns, clicks and networks.
Next, those users might see a Dyson vacuum, a pair of Forzieri wingtips or a single woman with a six-figure income living within a ten-mile radius. In fact, that's where Levchin thinks the first real opportunity lies - hooking up users with like-minded people. If this all sounds vaguely creepy, Levchin is careful to say he's rolling out features slowly and will only go as far as his users will allow.
But he sees what many others claim to see: Most consumers seem perfectly willing to trade preference data for insight. As Levchin and his charges set out to build a new type of search engine - a new Google - one question becomes obvious: Where's the old Google? VP of engineering Udi Manber refuses to comment on whether there's a recommender system in the pipeline. But it's a safe bet something is coming. Google's director of research, Peter Norvig, is an advisor to CleverSet, a recommender company in Seattle.
Kuhlke and Geitgey never had Google-sized ambitions for Whattorent. The co-founders, who recently took day jobs as software engineers on opposite ends of the country, didn't expect their site to be a huge moneymaker.
The firm's outlook for a more hawkish Fed comes just a few days ahead of key inflation readings this week that are expected to show prices rising at their fastest pace in nearly 40 years. If the Dow Jones estimate of 7. That figure is due out Wednesday. At the same time, Hatzius and other economists do not expect the Fed to be deterred by declining job growth.
Nonfarm payrolls rose by , in December , well below the , estimate and the second month in a row of a report that was well below consensus. However, the unemployment rate fell to 3. It's no easy feat, but if a company can pull it off and make the formula portable so it works on your mobile phone - well, such a tool could change not just marketing, but all of commerce.
Amazon Charts realized early on how powerful a recommender system could be and to this day remains the prime example.
The company uses a series of collaborative filtering algorithms to compare your purchasing patterns with everyone else's and thus narrow a vast inventory to just the stuff it predicts you'll buy. Its system favors popular, obvious items and tends to come off less like a trusted shopkeeper than a pushy salesman: If you liked the novel "The Corrections," you'll see suggestions to buy "The Discomfort Zone" and everything else Jonathan Franzen has written.
If you bought a gift for a baby shower, you're bound to get a stream of recommendations for cheap plastic toys and birthing blankets.
The new generation of recommenders will do better. Some employ filters that factor in more variables. Others analyze the contents of what they're recommending to grasp why you like something. A third category, hybrid recommenders, combines both strategies. Five weeks later, 37 registrants had already posted improvements to the Netflix system.
Two contestants were just short of halfway to the goal. Back at the Vortex, Kuhlke and Geitgey discuss how their parlor trick works. They'll draw on their knowledge of cinema and their experience categorizing hundreds of films - by star power, plot complexity, etc. Geitgey goes first. He scans the crowd and homes in on a target, a guy delivering meals and clearing dishes. On their site, Kuhlke and Geitgey use a series of odd questions to determine a user's personality.
This is what Geitgey sees: tattered jeans, a steel bracelet, a few tattoos. The target may be gathering dishes but seems too authoritative to be a busboy. He appears to be in his late 20s and is working, Geitgey surmises, "in a youth-trendy restaurant in the part of the city where people that age who don't have real jobs hang out. Geitgey makes a few leaps. Which movie? In a few short years, Pandora has become the most efficient new-music discovery mechanism in history.
That's not saying much, really. Consider the alternatives: scouring magazines for reviews, flipping through albums in the record store, listening to radio stations all play the same songs. At Pandora. By rating songs and artists, you can refine the suggestions, allowing Pandora to create a truly personalized station. Unlike collaborative filtering engines, Pandora understands each song in its database.
Forty-five analysts, many with music degrees, rank 15, songs a month on characteristics to gain a detailed grasp of each. A former musician and film composer, founder Tim Westergren came up with the idea for Pandora while scoring movies for directors.
I developed a genome in my head, and would say, 'Okay, you like this song; do you like this one? Four million people now use Pandora the same way those directors used Westergren. The next tune I hear might be from We Are Scientists, a Brooklyn indie band with, Pandora determines, similar "electric-rock instrumentation, subtle use of vocal harmony, and minor-key tonality. It's safe to say that few consumers are searching for this band.
But on Pandora, it's only a few degrees of separation from Bloc Party. To draw a line from Bloc Party to Fire When Ready, the Music Genome Project combs through hundreds of thousands of songs and millions of pieces of user feedback. It's an impressive technological accomplishment but not nearly as impressive as the implications. If Pandora can nail me as a fan of a band that few people have ever heard of, and my musical tastes are an intimate expression of who I am, then Pandora could introduce me to a lot more than music.
Take it from Jason Rentfrow. So Rentfrow likes sophisticated jazz. What sort of picture does that paint? If you think he's intelligent, articulate, a bit geeky and soft-spoken and wears glasses, well, you'd be relying on stereotypes.
And you'd be right. Of course no person is one-dimensional. Rentfrow also likes the rap group A Tribe Called Quest. With that information, you would probably redraw his caricature. Now maybe he seems younger and more open-minded. Revealing a list of his 50 favorite artists would flesh him out even more.
0コメント