Could Google swing an election? April 3, 2013Posted by Cameron Shelley in : STV302 , comments closed
This item from PBS discusses research that indicates that the ranking that search engines assign to election news can affect the electoral intentions of voters. Here is the interview:
As Dr. Epstein points out, the claim is not that Google could dictate an election outcome, but that it could influence enough voters to make a crucial difference in how people intend to vote. Of course, this claim also assumes that the effect would resist other sources of influence, such as the voting intentions of voters’ friends and family, or what voters see on the evening news.
Dr. Epstein argues that strict regulation is in order, although what form the regulation would take is not clear. Search engines are, of course, in the business of ranking search results. We could not require them to order election news randomly, could we?
Another approach would be to allow search engines like Google to order such news in a way that is neutral. There are two obvious problems with this approach:
- How would you identify what constitutes election news? Would any news in which a politician or a political issue appear be fair game? That would take in quite a bit of information, much of it probably not truly electoral in nature. Identification could be narrowed by focus on items that describe elections explicitly. That approach might exclude quite a bit of relevant information, and could be easy to game by political hacks.
- What would a neutral (but not random) ranking system look like? Dr. Epstein’s research suggests one possibility. His experiment included lists sorted by pro-Abbott and pro-Gillard articles from the Australian federal election of 2010. A neutral list might be constructed by folding these lists together. However, we then need a quick way of rating news items according to their political slant, which may not be straightforward outside of an experimental setting.
Google might protest that it has no reason to distort search results in either way. Its interest is in results that users would like the most, whatever their political leanings might be. Such an approach may strengthen users’ biases, but it would not alter them, which is what Dr. Epstein is concerned about.
It might be interesting to find out if Dr. Epstein’s results apply to people’s preferences in general. That is, could Google influence people’s tastes in clothes, food, or cars, for example? I guess that is the job of Google ad services, like AdWords. Perhaps, then, in the interests of neutrality, we should regulate how political parties buy their ads.
TV watches you! March 21, 2013Posted by Cameron Shelley in : STV302 , comments closed
In America, you watch television.
In Soviet Russia, television watch you.
Soon, America will become more like the old Soviet Union. FastCompany notes that Samsung is set to offer a television that tracks the viewing habits of its users, in order to make entertainment recommendations. Google and Panasonic have already announced similar plans for their upcoming TV offerings.
Along with their resolution, the level of surveillance offered by these screens will be quite high, which will be of interest to advertisers:
Rob Enderle, an analyst and consultant with Enderle Group, said this model will become the norm as television gravitates to Internet platforms.
“Increasingly, TVs will know who is watching them and I expect advertisers will know shortly thereafter. This should result in shows and commercials you like more and even better products, but far less privacy.”
Stu Lipoff, a fellow at the Institute for Electrical and Electronics Engineers, said TV on mobile devices will have similar characteristics, with considerable amounts of data which can be gleaned about viewers.
Of course, millions of Xbox users have already agreed to this sort of arrangement, so it is not completely new.
Still, does it sound creepy or Orwellian? Unlike the novel 1984, users will be given options regarding the level of privacy that they want:
[Chinese manufacturer] TCL’s [Haohong] Wang says, meanwhile, the TV makers are not interested in tracking people and will allow them options.
“We are an equipment company. What we want is to give a good user experience,” he said. And if viewers feel uncomfortable with being monitored they don’t have to use those features, he said: “They can just turn it off.”
In other words, the default will be full surveillance. The alternative will be something less. Nice.
(Lali Masriera/Wikimedia commons)
Google glasses February 23, 2012Posted by Cameron Shelley in : STV302 , comments closed
The New York Times reports that Google is designing a pair of glasses that will display your smartphone information for you right in front of your eyes. Basically, the glasses will take the place of your smartphone and use the lenses of your glasses as their display. Naturally, the glasses will run on Android and will be priced comparably to an Android smartphone.
(Erik Möller/Wikimedia commons)
So, the phone will be able to do those things that your smartphone now does, e.g., display text messages, broadcast your position via Google Lattitude (say), and allow you to talk on the phone without holding a phone in your hands. Also, the device will have a camera that will allow the phone to double as an augmented reality device:
The glasses will have a low-resolution built-in camera that will be able to monitor the world in real time and overlay information about locations, surrounding buildings and friends who might be nearby, according to the Google employees.
Reminds me a bit of the police camera-glasses I discussed yesterday.
Naturally, Google is concerned about the privacy implications that could be raised here. (I will wait until you finish laughing. Ready now?)
Internally, the Google X team has been actively discussing the privacy implications of the glasses and the company wants to ensure that people know if they are being recorded by someone wearing a pair of glasses with a built-in camera.
I guess that the glasses could come with a little LED that blinks as the wearer records everyone. Or perhaps the glasses could display a map of other wearers of the glasses in the vicinity. Or people could log in to their Google accounts to request that their faces be blurred in any recordings, a la Streetview. In any event, as long as you know you are being recorded, your privacy is assured!
Should we not also be concerned about how such glasses will distract people as they walk around town or drive their cars? It might be legal to drive with these glasses on in Ontario, where only hand-held devices are forbidden to drivers, but it hardly seems like a good idea.
In any event, the greatest advantage of the new Glasses will be in the increased efficiency they give to users. Facial recognition software will allow the glasses to recognize and identify my friends (and enemies) for me, freeing up the facial recognition part of my cortex for more important jobs, like monitoring my Google+ feed. When the display resolution gets high enough, I will not actually need to go anywhere. Instead, I can log in to Virtual Streetview and meet avatars in virtual reality, while beholding Google’s virtual ads.
Perhaps these worries are the products of a hyperactive imagination. I hope so. However, it seems that Google wants to build a wall around the users of their goods and services, so that they can be submerged in Google juice and pumped for cash as efficiently as possible.
Your google record October 24, 2011Posted by Cameron Shelley in : STV302 , comments closed
Our Internet searches certainly say something about us. That something is not always admirable.
Consider the news that Mohammad Shafia of Montreal and his sons Tooba and Hamed are accused of the “honour killings” of his first wife Rona and their daughters Zainab, Sahar, and Geeti. Allegedly, Mr. Shafia felt that his daughters had shamed him and the family through licentious conduct. The method chosen to restore family honour was to kill the daughters and first wife, a traditional practice in some parts of the world.
No, Google did not make him do it. Instead, police have turned up incriminating evidence by searching the laptop that they seized from his son:
Forensic experts found evidence of a Google search 10 days before the deaths.
“The words entered were: ‘Where to commit a murder,’ ” Lacelle said. A few weeks earlier there had been a search on the computer for, “Can a prisoner have control over their real estate,” the prosecutor said.
Of course, such evidence may be taken out of context. Someone might read about a murder in a news article and then enter a similar search in order to find out more about the topic. Even so, it does not bode well for Mr. Shafia.
There is currently a similar situation before courts in the UK. There, a Dutch engineer named Vincent Tabak is suspected in the death of Joanna Yeates. He also has been found to have carried out suspicious Google searches prior to the event:
Among the phrases Tabak Googled were “sexual offence explained” and “definition of sexual assault”, the jury was told.
Next day he looked at online maps and images of Longwood Lane, the road three miles from her Bristol flat where her body was discovered.
Later, the jury was told, he researched subjects including: “How does forensic identification work?” and the location of CCTV cameras in Canynge Road, Clifton, where Tabak and Yeates lived.
He researched “body decomposition time” and an article about a man who strangled his wife and pleaded diminished responsibility.
Perhaps Mr. Tabak was cooking up a story in case he was questioned by police.
These incidents display the trade-offs of services such as Google. I take it we are all familiar with the advantages brought by the access to knowledge that Google provides. Not all of that information works to the good of society, such as information about how to commit crimes, or where. Of course, the default of recording search terms does help police to gather relevant evidence after a crime has been detected.
Does Google have any obligations to society in cases like these? Should it, for example, censor search results to prevent potential criminals from finding information useful to them? Such censorship would interfere with rights to freedom of expression and would likely hamper public discourse on the topic of murder or other controversial subjects. Perhaps Google could be required to flag suspicious search histories. Banks are required to flag transaction histories that seem suspicious. It was apparently a record of suspicious transactions that brought former NY governor Elliot Spitzer to the attention of police. Should a similar standard be applied to Google? Perhaps, although the problem of dealing with false positives looms large: Any writer doing research for a murder mystery or even simple public discussion of prominent crimes would bring suspicion on many innocent people. In the end, the default of recording search histories in browsers may be the best approach available.
Turnabout is fair play? March 9, 2011Posted by Cameron Shelley in : STV302 , comments closed
The Internet and smartphones are playing an increasing role in mediating relations between citizens and politicians. For example, there are smartphone apps that allow citizens to complain to city hall about potholes, bus schedules, police behavior, etc.
From my home town of Guelph, Ontario, comes an interesting variation on this relationship. Specifically, it seems that one of our councilors, Maggie Laidlaw, uses the ‘net in order to characterize complaints that she receives from citizens. For example, she uses Google Maps to investigate people who complain about taxes and city spending:
“When you happen to ‘Google Map’ those addresses, they are all living in pretty nice houses,” sometimes with a boat as well as a number of autos, she said.
“I don’t pay much attention to those emails anymore,” she told council.
An interesting practice. I wonder what other creative uses politicians make of the ‘net, besides PR on their blogs and so on?
Outsourcing university email April 13, 2010Posted by Cameron Shelley in : STV302 , comments closed
A recent NPR news story explains that some universities are outsourcing their IT to Google. That is, students use Gmail and Google Tools to communicate and work on assignments, all organized through a portal dedicated to students from that university.
The service is provided to Wesleyan at no charge, which is a big help to the University’s IT budget. Also, students find the Google services effective and easy to use, whereas the University’s own services were considered “clunky”.
Google claims that there are no catches to the service. Although they are in possession of students’ data, they will not scan it nor will they charge for it in future. Thus, privacy concerns are minimized. Of course, security concerns remain (Gmail has been hacked in the past) but such concerns would apply anywhere. And the portal will not display ads, and so concerns that the university experience would resemble a shopping mall are diminished. The only reward Google seems to want is to burnish their good name, i.e., brand, with university students.
Would you like to see Waterloo adopt Google services to provide IT facilities to its students?