Fuzz: An anti-"big data" streaming media company

According to foreign media reports, a study published in Friday by Future tense, a foreign publication that specializes in new technologies that can affect human society, policy and culture, says that algorithms and "big Data" do make it easy to figure out what people like, but they also stifle creativity.

The comments are as follows:

Of all the startups that were founded last year, Fuzz is certainly one of the most alluring companies that most people overlook. Fuzz claims to be a "human radio station with no robotics at all", and fuzz has challenged this trend as people increasingly trust algorithms in discovering new music. Fuzz praised the important role played by the artificial DJ, where the artificial DJ refers to a group of fixed users of Fuzz, who were invited to upload their music to the Fuzz Web site to create and share their own radio stations.

The idea behind fuzz or the hope is that the artificial DJ can pass the information that the algorithm cannot deliver. It wants to move in the opposite direction to Pandora, which is mainly doing all the heavy work through various algorithms. "Users have a huge demand for this type of assistant experience, and we just want to convince people that the most compelling recommendation comes from living people," said Jeff Yasuda, founder of Fuzz, in a Bloomberg interview last September. ”

However, although the fuzz on the line has hardly aroused anyone's attention, but the various algorithms in the art process plays an important role has become more and more can not be ignored. Andreul-les-Annards Andrew Leonard, a technical critic at Sharon, also highlighted the important role of the algorithm in a commentary he wrote about Netflix's service to the original show--house of Cards. The secret behind the originality of those shows is now known to all, that Netflix, first of all, studied user records and found that the redesign of a British TV series with the same name could be a great success, especially with Kevinspethy Kevin Spacey and by David Fernchell David Fincher) to direct the re-production of the TV play.

"Is this famous director still able to survive in an age when this computer algorithm has become the ultimate concern?" Leonard. "Netflix collects a lot of user data, such as how many pause buttons a user clicks on the first season of a TV show, and so on, and he wants to know what impact Netflix's collection of data will have on future TV dramas," he said.

Many other industries face similar problems. Amazon, for example, collects a lot of information about the user's reading habits through its Kindle E-book reader, including which books the user has read? What books have you not read? Which chapters do they generally prefer to skip? Which chapters are most closely watched? How often do they have to look up a dictionary and underline the paragraphs below? (not only do Amazon companies do this, but other E-book readers are collecting similar data.) ）

Using this data, Amazon can predict all the elements needed to get readers to read a book at a gulp. Perhaps Amazon can even provide readers with other outcomes that will make readers happier. As the latest report on the future of the entertainment industry suggests, we are in a world where many things can be adapted to create a more attractive and interactive future.

With all the user data available, it would be foolish for Netflix to get out of the movie-making business. Like Netflix, Amazon has found this, so it must enter the publishing industry. Amazon, however, is actually deeper than Netflix, because it also runs a book-selling website that knows all the buying behavior of consumers and what prices consumers are willing to pay. Now, Amazon operates 6 electronic publications, and it intends to add more publications.

The music industry accepted a similar approach a few years ago, collecting and building a huge database of previously popular songs and failed songs, and using it to predict whether new songs could become popular songs. The advantage of this approach is obvious: new artists need not have a huge network to contract with the record company, and in the past, networking is one of the prerequisites for the success of new artists. Now, new artists just use the past success data to deduce a song, it is likely to become a popular song.

But the disadvantage of this approach is clear: the new songs we can finally hear may sound similar, lacking in originality and vigor. As Kristofstein Christopher Steiner in his new book, this automation (automate this): "Such technology may bring us new entertainers, but because their judgments are based entirely on past popular songs, So the new songs we hear may be the same type of fast food as the ones we've forgotten. This is clearly a weakness of this technology. ”

The Watson supercomputer developed by IBM will be used to handle thousands of legal and medical documents in order to provide support for people to make various decisions. Because there are too many documents to read, no lawyer or member of the Institute can see them. If the goal is simply to analyze what was sold in the past and use it as a predictor of possible future selling, Watson can easily be applied to music, film and book Fields.

However, while doing so helps improve sales, it may also stifle cultural innovation. How could Watson predict the rise of Impressionist paintings, futuristic poetry or new wave movies? How could it agree with Stravinsky (Stravinsky)? And big data is likely to Miss Dada.

To understand the limitations and opportunities offered by the algorithm for artistic creation, we need to understand that the opportunity provided by the algorithm is usually composed of 3 elements, namely discovery, production, and recommendation. Startups like Fuzz aim for the third element, which is to suggest that some users want to provide wizards for themselves by living people rather than algorithms.

Fivebooks, which provides book referral services for readers who love reading, also implements a model similar to fuzz, because it believes that living people can do better than rigid algorithms in recommending books. Amazon has done a good job in book recommendation, but Fivebooks has recommended to readers the works of Paul Kugman (Paul Krugman), Haroldbrum (Harold Bloom) and Inmekivan (Ian McEwan) in a different measure. In fact, manual referrals and algorithm recommendations can exist at the same time, at least for the foreseeable future, because readers will find a balance between the two recommended models.

But when it comes to finding new talent and studying the direction of future creation, the algorithm is less optimistic. After all, the algorithm recommendation makes sense only if there is really great work. If the algorithm chooses a work based on previous successes and timely feedback from the reader, the sales of new works may increase, but will not provide more valuable benefits to the industry.

The first signs are not encouraging. Last December, the English-language Global Times published a report on the local punk band Bear Warrior, which found an ingenious way to detect the response of listeners to their songs. The lead singer is a graduate student of the precision instrument speciality of Peking University, he designed a device called "pogo thermometer" that could detect the strength of the audience's dance steps through a series of sensors installed in a concert hall rug, then send the signal to a central computer, and finally let the central computer analyze the signal, Help the band improve their deductive style.

According to the Global Times, the band found that fans would start swinging when the drums rang, and when the lead singer sang to the top of the song, the fans would reach the climax of the dance. As the lead singer put it: "This data can help us understand how we can improve our deductive approach, and let the audience respond to our music work as we would like to see it." ”

Maybe it does help to improve their deductive style, but when did punk music become so nuanced? Making the audience happy is what management consultants need to think about, but it's not something that punk musicians should be thinking about or even obsessed with! When the Sex Pistol band (Sex pistols) sings, the carpet in the concert hall has only one function, that is, for the fans to dance on it and never install any sensors. But the sexy pistol band created a revolution in punk music, and the Bear Band, at best, turned punk music into a career for them to earn a living.

(Responsible editor: The good of the Legacy)