Neural Networks Finally Have Their Day
May 11, 2015
The Toronto Star offers a thoughtful piece about deep learning titled, “How a Toronto Professor’s Research Revolutionized Artificial Intelligence.” Professor Geoffrey Hinton was instrumental in pursuing the development of neural network-based AI since long before the concept was popular. Lately, though, this “deep learning” approach has taken off, launching many a product, corporate division, and startup. Reporter Kate Allen reveals who we can credit for leading neural networks through the shadows of doubt:
“Ask anyone in machine learning what kept neural network research alive and they will probably mention one or all of these three names: Geoffrey Hinton, fellow Canadian Yoshua Bengio and Yann LeCun, of Facebook and New York University.
“But if you ask these three people what kept neural network research alive, they are likely to cite CIFAR, the Canadian Institute for Advanced Research. The organization creates research programs shaped around ambitious topics. Its funding, drawn from both public and private sources, frees scientists to spend more time tackling those questions, and draws experts from different disciplines together to collaborate.”
Hooray for CIFAR! The detailed article describes what gives deep learning the edge, explains why “machine learning” is a better term than “AI”, and gives several examples of ways deep learning is being used today, including Hinton’s current work at Google and the University of Toronto. Allen also traces the history of the neural network from its conceptualization in 1958 by Frank Rosenblatt, through an era of skepticism, to its recent warm embrace by the AI field. I recommend interested parties check out the full article. We’re reminded:
“In 2006, Hinton and a PhD student, Ruslan Salakhutdinov, published two papers that demonstrated how very large neural networks, once too slow to be effective, could work much more quickly than before. The new nets had more layers of computation: they were ‘deep,’ hence the method’s rebranding as deep learning. And when researchers began throwing huge data sets at them, and combining them with new and powerful graphics processing units originally built for video games, the systems began beating traditional machine learning systems that had been tweaked for decades. Neural nets were back.”
What detailed discussion of machine learning would be complete without a nod to concerns that we develop AI at our peril? Allen takes some time to sketch out both sides of that debate, and summarizes:
“Some in the field believe that artificial intelligence will augment, not replace: algorithms will free us from rote tasks like memorizing reams of legal precedents and allow us to pursue the higher-order thinking our massive brains are capable of. Others think the only tasks machines can’t do better are creative ones.”
I suppose the answers to those debates will present themselves eventually. Personally, I’m more excited than scared by the possibilities. How about you, dear reader?
Cynthia Murrell, May 11, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Digital Economy Growth Engines Lose RPMs
April 24, 2015
Short honk: I read several articles about the financial reports of Facebook, Google, and Yahoo. I enjoyed the explanations about the revenues and profits. Here are the write ups open on my desktop monitor at this moment:
- “Despite Headwinds, Analysts See Even Larger Facebook Upside Into 2016”
- “Google Caps Costs as Growth Slows” for which you may have to pay to read.
- “Yahoo Q1 Results Miss Expectations on Both Lines”
Is there a message to be decrypted from these data? Yep.
Stephen E Arnold, April 24, 2015
Google Allegedly Just Gets a Good Idea: A Data Platform
April 24, 2015
Wow. I read some interesting and often crazy stuff. But this is a keeper. Navigate to “Google Builds a Data Platform That’s the Last Piece of Its Ad Empire. Connects Dots for Marketers and Challenges Facebook.” Never mind that the Google has been working on the data platform thing for advertising for what is it now, 12 or 13 years. Never mind that the guts of the ad system’s interfaces have been a project at the Google for more than a decade. Never mind that the guts of the data platform idea originated before Google hired Drs. Halevy and Guha along with hundreds of other scientists and engineers eager to knit together data from Google’s various repositories. But, hey, it is an advertising Web site, and I assume advertising experts are a heck of a lot more informed than little old me.
I read:
Of course, Google faces regulatory scrutiny for any move it makes, as well as talk of anti-competitive practices. In fact, the company was charged in Europe last week with behaving like a monopoly in search. The ad tech community has been concerned that Google is offering all the services that lock advertisers into its ecosystem and squeeze out rivals.
What the write is about is the “lead” which Facebook has over Google. The problem is not technology, in my humble opinion. The problem is that Google is focused on technology and Facebook was built to allow a person to get a date. Facebook followed its social-human thing, and the GOOG has been embracing the ever lovable zeros and ones. There are Googlers at Facebook, but Facebook will not become a Google. I would argue that Google cannot become a Facebook.
The data platform is secondary to the source of the information fueling the respective systems. Facebook users are the content sources. Google’s content comes from other places. Both companies face significant challenges and neither is likely to morph into another.
Why not merge into a Googbook or Facegle? If it works for Comcast and Time Warner, it might work for Google and Facebook. Ad buys just become easier. Ad people often prefer the easy approach.
Stephen E Arnold, April 24, 2015
The Elusive Video Recognition
April 22, 2015
Pictures and video still remain a challenge for companies like Google, Facebook, Apple, and more. These companies want to be able to have an algorithm pick up on the video or picture’s content without relying on tags or a description. The reasons are that tags are sometimes vague or downright incorrect about the content. VentureBeat reports that Google has invested a lot of funds and energy in a deep learning AI. The article is called “Watch Google’s Latest Deep Learning System Recognize Sports In YouTube Clips.”
The AI is park of a neural network that is constantly fed data and programmed to make predictions off the received content. Google’s researchers fed their AI consists of a convolutional neural network and it was tasked with watching sports videos to learn how to recognize objects and motions.
The researchers learned something and wrote a paper about it:
“ ‘We conclude by observing that although very different in concept, the max-pooling and the recurrent neural network methods perform similarly when using both images and optical flow,’ Google software engineers George Toderici and Sudheendra Vijayanarasimhan wrote in a blog post today on their work, which will be presented at the Computer Vision and Pattern Recognition conference in Boston in June.”
In short, Google is on its way to making video and images recognizable with neural networks. Can it tell the differences between colors, animals, people, gender, and activities yet?
Whitney Grace, April 22, 2015
Stephen E Arnold, Publisher of CyberOSINT at www.xenky.com
A Former Googler Reflects
April 10, 2015
After a year away from Google, blogger and former Googler Tim Bray (now at Amazon) reflects on what he does and does not miss about the company in his post, “Google + 1yr.” Anyone who follows his blog, ongoing, knows Bray has been outspoken about some of his problems with his former employer: First, he really dislikes “highly-overprivileged” Silicon Valley and its surrounds, where Google is based. Secondly, he found it unsettling to never communicate with the “actual customers paying the bills,” the advertisers.
What does Bray miss about Google? Their advanced bug tracking system tops the list, followed closely by the slick and efficient, highly collaborative internal apps deployment. He was also pretty keen on being paid partially in Google stock between 2010 and 2014. The food on campus is everything it’s cracked up to be, he admits, but as a remote worker, he rarely got to sample it.
It was a passage in Bray’s “neutral” section that most caught my eye, though. He writes:
“The number one popular gripe against Google is that they’re watching everything we do online and using it to monetize us. That one doesn’t bother me in the slightest. The services are free so someone’s gotta pay the rent, and that’s the advertisers.
“Are you worried about Google (or Facebook or Twitter or your telephone company or Microsoft or Amazon) misusing the data they collect? That’s perfectly reasonable. And it’s also a policy problem, nothing to do with technology; the solutions lie in the domains of politics and law.
“I’m actually pretty optimistic that existing legislation and common law might suffice to whack anyone who really went off the rails in this domain.
“Also, I have trouble getting exercised about it when we’re facing a wave of horrible, toxic, pervasive privacy attacks from abusive governments and actual criminals.”
Everything is relative, I suppose. Still, I think it understandable for non-insiders to remain a leery about these companies’ data habits. After all, the distinction between “abusive government” and businesses is not always so clear these days.
Cynthia Murrell, April 10, 2015
Stephen E Arnold, Publisher of CyberOSINT at www.xenky.com
Facebook Users Lack Understanding of Filters: No Big Surprise
March 29, 2015
Let me be clear. I am not a Facebook user. One of the goslings configured the Beyond Search blog to send content to a Facebook page. I, however, do not need a stream of information about my high school and college classmates. At my last reunion, the 50th, I saw only two mobile phones: My wife’s and mine. Obviously central Illinois is not a technology hot spot for the over 70 set.
I read “Many, Many Facebook Users Still Don’t Know That Their News Fees Are Filtered by an Algorithm.” Big whoop. Most of the MBAs I know are clueless about Google’s personalization functions and don’t have much appetite for understanding that what you see may not be what is available. For these cohorts, a little learning is just fine. Drinking from a spring is okay as long as the water comes from an authentic source like Dasani. Isn’t that Coca Cola’s outfit?
The write up reveals what strikes me as a no brainer type factoid:
But a majority of everyday Facebook users in a recent study had no idea that Facebook constructs their experience, pushing certain posts into their stream and leaving others out. And worse, many participants blamed themselves, not Facebook’s software, when friends or family disappeared from their news feeds.
The article reports:
While some participants were upset by the idea that Facebook was changing their social experience, more than half of the study participants “came to appreciate the algorithm over the course of the study.” Most came to think that the filtering and ranking software was actually doing a decent job. “Honestly I have nothing to change which I’m surprised!” one said. “Because I came in like ‘Ah, they’re screwing it all!’”
Sigh. Is there a remedy for this lack of understanding? Nope.
Do most online “experts” care? Nah, but some of them charge windmills with their iPad Airs as a shield.
The reality is that a comprehensive understanding of a particular content domain requires good, old fashioned research. The idea is to read, talk to informed individuals, gather additional primary data, analyze what you collect, and then figure out who knows what about a topic.
We are doing this type of grunt work about one facet of the Dark Web. Early results are in. Most of the people writing about the Dark Web are not doing a particularly good job of explaining where the “dark” content lives, how to find it, or what the content reveals about a fundamental shift in online usage for a small but important and interesting group of users worldwide.
If one cannot understand what Facebook is doing, the Dark Web is of zero consequence. If a Google user accepts search results as objective, I am not sure there is much hope for remedial intervention.
Net net: At a time when ease, convenience, short cuts, and distractions are of primary importance, thinking about information is not of much interest to many people.
“Hey, after the NCAA games, let’s binge watch Breaking Bad. We can post our comments on Facebook too!”
Sound fun? Oh, wait. I have to take this call, send an SMS, and post a picture of our pizza to Facebook. Cool.
Stephen E Arnold, March 29, 2015
Painting an IT Worker’s House Requires an NDA
March 27, 2015
You would not think that contractors, gardeners, painters, plumbers, and electricians would have to sign an non-disclosure agreement before working on someone’s home, but according to the New York Times it is happening all over Silicon Valley. “For Tech Titans, Sharing Has Its Limits” explains how home and garden maintenance workers now have to sign NDAs for big name tech workers just like they have to with celebrities. Most of the time, workers do not even know who they are working for or recognize the names. This has made it hard to gather information on how many people require NDAs, but Mark Zuckerberg recently had a lawsuit that sheds some light about why they are being used. He goes to great lengths to protect his privacy, but ironically tech people who use NDAs are the ones who make a profit off personal information disclosures.
“The lawsuit against Mr. Zuckerberg involves a different residence, 35 miles south in Palo Alto. In it, a part-time developer named Mircea Voskerician claims that he had a contract to buy a $4.8 million house adjoining Mr. Zuckerberg’s residence, and offered to sell a piece of the property to Mr. Zuckerberg. He says that in a meeting at Facebook headquarters in Menlo Park, he discussed a deal to sell his interest in the entire property to Mr. Zuckerberg. In exchange, he says, Mr. Zuckerberg would make introductions between him and powerful people in Silicon Valley, potential future business partners and clients. Mr. Voskerician passed up a better offer on the house, the suit contends, but Mr. Zuckerberg did not follow through on the pledge to make introductions.”
Voskerician said he only signed the NDA on as a condition to the proposed agreement, but Zuckerberg’s legal representation says the NDA means all information related to him. On related terms, Facebook is making more privacy rules so only certain people can see user information. It still does not change how big name IT workers want their own information kept private. It seems sharing is good as long as it is done according to a powerful company’s definition of sharing.
Whitney Grace, March 27, 2015
Stephen E Arnold, Publisher of CyberOSINT at www.xenky.com
Study Find Millennials Willing to Pay for News to a Point
March 26, 2015
The article titled Millennials Say Keeping Up With the News Is Important To Them—But Good Luck Getting Them To Pay For It on NiemanLab explores the findings of a recent study by the Media Insight Project in partnership with the American Press Institute. A great deal of respondents get their news from Facebook, although the majority (88%) said it was only occasionally. Twitter and Reddit also made the list. Interestingly, millennials claimed multiple access methods to news categories across the board. The article states,
“The survey asked respondents how they accessed 24 different news topics, from national politics and government to style, beauty, and fashion. Facebook was either the number one or two source of information for 20 of the 24 topics, and in nine of those topics it was the only source cited by a majority of respondents. Search was the second most popular source of information, ranking first or second in 13 of the 24 news topics.”
In spite of the title of the article, most millennials in the study were willing to pay for at least one subscription, either digital or print. The article doesn’t mention the number of people involved in the study, but deeper interviews were held with 23 millennials, which is the basis for the assumptions about broader unwillingness to pay for the news, whether out of entitlement or a belief that access to free news is a fundamental pillar of democracy.
Chelsea Kerwin, March 26, 2015
Stephen E Arnold, Publisher of CyberOSINT at www.xenky.com
Facebook Gains Natural Language Capacity with Via AI Acquisition
February 11, 2015
Facebook is making inroads into the natural language space, we learn from “Facebook Buys Wit.ai, Adds Natural Language Knowhow” at ZDNet. Reporter Larry Dignan tells us the social-media giant gained more than 6,000 developers in the deal with the startup, who has created an open-source natural language platform with an eye to the “Internet of Things.” He writes:
“Wit.ai is an early stage startup that in October raised $3 million in seed financing with Andreessen Horowitz as the lead investor. Wit.ai aims to create a natural language platform that’s open sourced and distributed. Terms of the deal weren’t disclosed, but indicates what Facebook is thinking. As the social network is increasingly mobile, it will need natural language algorithms and knowhow to add key features. Rival Google has built in a bevy of natural language tools into Android and Apple has its Siri personal assistant.”
Though the Wit.ai platform is free for open data projects, it earns its keep through commercial instances and queries-per-day charges. Wit.ai launched in October 2013, and is headquartered in Palo Alto, California.
Cynthia Murrell, February 11, 2015
Sponsored by ArnoldIT.com, developer of Augmentext
Is Facebook the Internet?
February 9, 2015
There has been a buzz in some circles that Facebook is the Internet. I think that one’s point of view plays a part in feeling comfortable with the statement. The author of “Millions of Facebook Users Have No Idea They’re Using the Internet” finds the idea more than a little intriguing. The write up makes reference to Facebook users far from Silicon Valley.
I highlighted this passage:
Since at least 2013, Facebook has been making noises about connecting the entire world to the internet. But even Sheryl Sandberg, Facebook’s operations head, admits that there are Facebook users who don’t know they’re on the internet. So is Facebook succeeding in its goal if the people it is connecting have no idea they are using the internet? And what does it mean if masses of first-time adopters come online not via the open web, but the closed, proprietary network where they must play by Facebook CEO Mark Zuckerberg’s rules?
The write up points out that there is a slip twixt cup and lip. However, the theme of Facebook is the Internet continues to form the spine of the write up. I noted this passage:
Already services are starting to move away from the open web and to Facebook. And it’s happening not just in the poor world, but in poor parts of the developed world, where there also exists a sense among some that using an app isn’t the same as using the internet, which requires a web browser like Safari or Internet Explorer.
My view is that the notion of the open Internet is going to be a thorny issue. Governments are clamping down on some types of Internet sites. I checked one extremist Web site based in France. On February 5, 2015, the site was online. On February 6, 2015, the site returned 404s. However, some of the somewhat disturbing videos posted by the Web site remaining available on YouTube.com.
The idea for a state-certified information service may have some appeal. Did you explore Sputnik? Have you encountered issues with site access in China, Iran, or Turkey?
Is there a future in walled gardens?
Stephen E Arnold, February 9, 2015
-
- Subscribe to Beyond Search
Feature archive
News archive
-
