Link to the University of Waterloo home page

Link to the Centre for Society, Technology and Values home page

jump to navigation

Buridan’s robo-ass September 16, 2014

Posted by Cameron Shelley in : STV202, STV302 , add a comment

There is an interesting piece in New Scientist about an “ethical trap” for robots set by researchers at the Bristol Robotics Laboratory. In the trap, one robot is given the job of preventing others from falling into a hole by intercepting them. It seems to be programmed to save the nearest robot to it first. This works fine until it faces a pair of robots, both heading towards the hole at the same time and at an equal distance from their would-be rescuer. In that case, the rescue robot seems to hesitate about what to do, leaving both of the others to fall into the hole.

The problem illustrates an issue for programmers of all sorts of robots: how moral decisions should be made by our gear. (By moral decision, I mean a situation in which several courses of action are possible at least some of which may be considered moral or immoral.) Designers of military drones, for example, are already concerned about the need for “ethical governors”, and self-driving vehicle designers are facing similar issues.

The dilemma faced by the rescue robot is not unprecedented. It is reminiscent of an ancient problem known by the name of “Buridan’s ass“. In a nutshell, Buridan (a 14th Century French philosopher) argued that being moral means always choosing the best alternative course of action available. His critics satirized this idea with a thought experiment: What if a hungry ass were placed equidistant from two bales of hay? Since neither alternative is better than the other, the ass would remain motionless and starve. How absurd!

The hesitating robot is a kind of Buridan’s robo-ass: faced with two equally good options, it cannot decide and thus loses the opportunity to pursue either one.

How would you resolve this problem, and what other consequences would your solution have?

As our gear becomes more autonomous and as we rely on it more implicitly, such situations will increasingly move from the realm of philosophical puzzles to practical problems.

Hello selfie! September 15, 2014

Posted by Cameron Shelley in : STV202, STV302 , add a comment

The selfie is here to stay. However, there remains controversy about when and where they should be taken. Here are some recent instances of this conundrum:

So, keep pointing that camera backward; it may be worth a fortune someday!

Fonts in the news September 11, 2014

Posted by Cameron Shelley in : Uncategorized , add a comment

Some people are nuts about fonts while others rarely spare them a thought. (For whatever it is worth, I am in the latter camp.) However, the design and application of fonts do occasionally make news in the wider world. Three interesting, font-related stories have come up recently that seemed worth sharing:

One point about Comic Sans is that it has no kerning, which was not commonly available in 1994 when the font was propagated with Microsoft Windows. So, it would help solve the kerning problem with certain words on billboards. Of course, it might attract adverse comment for other reasons.

From DNN headquarters September 10, 2014

Posted by Cameron Shelley in : STV202, STV203 , add a comment

From the world headquarters of the Drone News Network:

Girl drinks September 8, 2014

Posted by Cameron Shelley in : STV202 , comments closed

Gender is an important cultural distinction. Although it relates to biological differences between the sexes, it is elaborated significantly in the designed environment. Things are sometimes designed differently for men and women. How such differences arise is an interesting matter. The adage “shrink it and pink it” exemplifies one common approach: It takes designs for men as the default and suggests miniaturization and change in palette. Women do tend to be smaller than men and may find smaller items, e.g., chairs and power tools, more comfortable. Preference for pink seems more like a cultural matter.

Gender can also be a significant factor in packaging design. Consider recent efforts to package and market hard cider as a man’s drink. Cider was a popular drink for men in North America during the 19th Century. One of the consequences of the spread of apple trees in the continent was a fondness for the beverage in rural areas. However, with increasing urbanization, purity laws, and the temperance movement, cider had fallen out of favor. By the 20th Century, beer became a canonical man’s drink. There is a “Joe six-pack” but not a “Jane six-pack” in American culture.

In order to get men back to cider drinking, producers are creating packages that portray its manliness:

As for Snake Bite, Barr and cofounder Tommy Hester took care in designing the packaging to stand out on shelves in a way that would appeal to men. Instead of a paper wrapper, the logo is painted directly onto the glass, with a sharp-toothed, yellow-eyed snake encircling the bottle. The result is more expensive to produce but also more tactile, giving the raised scales a sinister verisimilitude. “We went with more of an aggressive label because we wanted a cider that guys could drink,” Hester says.

Angular and poisonous, the brew is meant to be a hit with the XY chromosome crowd.

What this measure suggests is that there is no inherent gender in many goods, especially bulk goods such as beverages. Therefore, if gender distinctions are to be made, they must be imposed. Scholars who study such impositions call them “mythologies”, that is, cultural stories about the place that things have in the world. Girl drinks and boy drinks are one such mythology.

For the purpose of further illustration, I leave you with the following Kids in the Hall sketch, entitled “The Girl Drink Drunk”. See if you can spot the design qualities that differentiate girl drinks from boy drinks!

HI! I’M ELIZA. WHAT’S YOUR PROBLEM? September 5, 2014

Posted by Scott Campbell in : STV 210, STV205 , comments closed

Cam’s post yesterday reminded me of a program called ELIZA, a very simple Rogerian pseudo-psychotherapist created by famed computer scientist Joseph Weizenbaum in the mid 1960s.

Using almost no information about human thought or emotion, [the program] sometimes provided a startlingly human-like interaction… When the “patient” exceeded the very small knowledge base, [the program] might provide a generic response, for example, responding to “My head hurts” with “Why do you say your head hurts?” A possible response to “My mother hates me” would be “Who else in your family hates you?” ELIZA was implemented using simple pattern matching techniques, but was taken seriously by several of its users, even after Weizenbaum explained to them how it worked.

That last part is what struck me in relation to yesterday’s post about how people might be more inclined to share personal information with computers. There was a famous story at one point that after a session with ELIZA, Weizenbaum’s secretary refused to let him read the transcripts on the grounds of doctor-patient confidentiality!

Of course, don’t need a time-machine to try ELIZA today. There are many variations that can be found online or offline. I’m fond of the Gnu Emacs version myself, but I gave it a go online at http://www.masswerk.at/elizabot/ for fun. Here’s my transcript:

ELIZA: Please tell me what’s been bothering you.
YOU:   It’s pretty hot today
ELIZA: What does that suggest to you ?
YOU:   Climate change maybe?
ELIZA: Why the uncertain tone ?
YOU:   There are a lot of sceptics out there you know.
ELIZA: Are they always a lot of sceptics out there I know ?
YOU:   You don’t make much sense. Are you a real doctor?
ELIZA: What are your feelings now ?

Dialogue like this gets ridiculous pretty quickly, so the psychotherapeutic benefits seem questionable. But it remains an interesting early example that demonstrates the limits of Artificial Intelligence and the ability of humans to develop unjustified emotional affinity with their own machines.

Too thine own selfie be true September 5, 2014

Posted by Cameron Shelley in : STV302 , comments closed

As part of Warrior Black and Gold day (Sept. 6) on the UWaterloo campus, the Waterloo Warriors encourage people to wear the school colours and pose for their own camera. The photos can then be posted to the usual places with the hashtag “#mywarriorselfie”. A random selection of entries will be awarded some black-and-gold swag.

Although the Warriors are into selfies, not everyone is. The Queen, for example, declared to an interviewer that she finds them strange:

While the Queen has always been one to embrace modern technology, she has confided that she loathes the sea of mobile telephones that greets her every move in public.
Her Majesty told Matthew Barzun, the US ambassador, that she finds it “strange” to see nothing but the backs of mobile phones whenever she looks up.
“She was essentially saying: ‘I miss eye contact,’ ” Mr Barzun told Tatler magazine.

As a big celeb, it seems that the Queen meets more smart phones than people some days.

But do not despair. Selfies do not always distance people from one another. Consider the phenomenon of the “unselfie”, a self-portrait taken with a caption in support of a social cause. One example would be the “Bring back our girls” campaign in which people take a selfie with a suitable caption and post it with the hash tag “#BringBackOurGirls”, which I have mentioned earlier. However, an unselfie can be about any cause, such as cleaner air, cancer research, or ending poverty. The Huffington Post even has a whole news category devoted to this new form of expression.

Besides being exciting, strange, or inspiring, selfies can also be problematic. In case you have just emerged from hibernation, let me draw your attention to the story of how hackers stole explicit photos, including selfies, of female celebrities including Jennifer Lawrence and posted them to 4chan. The photos may have been taken from the iCloud accounts of these women using off-the-shelf software. The photos were clearly not meant for public consumption and appear to be just another way some male techies have of harassing women from their keyboards.

Moreover, some hackers have been supporting a campaign to get women to post nude unselfies of themselves with captions like “Solidarity with Jennifer Lawrence” and using the hashtag “#LeakforJlaw”. Happily, the ruse seems not to have worked. Posts using the hashtag display photos of things like leeks or leaky pipes.

It will be interesting to see what further surprises the selfie phenomenon holds in store.

Dr. Avatar September 4, 2014

Posted by Cameron Shelley in : STV202, STV302 , comments closed

In a recent posting, I discussed how students respond to robo-grading differently than to the conventional kind. To wit, students accept criticism more readily from the robo-grader because they do not take it personally. To put the matter another way, students find robo-grading to be technotonic and not technostressing.

Interestingly, a similar result has been found in another domain. Researchers Gale M. Lucas, Jonathan Gratch, Aisha King, and Louis-Philippe Morency found that people were more forthcoming with virtual therapists than with the real thing. The researchers were experimenting with the delivery of psychological therapy using an avatar. Subjects in the experiment all saw the same avatar. However, the group that were told that the avatar’s actions were controlled by a computer program more easily disclosed important personal information than did subjects who were told that the avatar’s actions were controlled by a human operator. (See the paper here.)

The explanation for both cases seems to be the same: People care less about what a computer thinks of them than what another human being does.

The immediate importance of the research lies in the suggestion that software therapists may be more effective than human therapists, at least for some tasks. Of course, the same techniques could be used by other programmers for other purposes. If Facebook is not happy with the level of your disclosure, for example, then it might employ virtual beings to prompt you to share more.

In what other ways might this result be put to work?

Three cheers for all the united workers of the world September 2, 2014

Posted by Scott Campbell in : STV 210, STV302 , comments closed

Yesterday was Labour Day. First, I hope everyone had a chance to celebrate the achievements of your fellow workers by relaxing and not working. More on that in a moment. Second, the day after Labour Day is traditionally the first day of school around here, and the beginning of Frosh Week at many Canadian universities which means, among many things, lots of loud chanting and singing.

Frosh week activities generally range from studious workshops designed to prepare students for the rigours of university life to pep-rallies where new students are taught school cheers to the more excessive and scandalous incidents involving underage drinking, rape-related chants, and secretive hazing rituals. These latter problems have quite rightly generated controversy and in response many universities have been rehabilitating frosh week, phasing out what had unthinkingly been accepted as traditional but were clearly inappropriate activities.

Setting aside these less pleasant problems, I wanted to highlight the more positive cheers, chants and songs that are a big part of frosh week and student life, and help create generate enthusiasm and maintain a collective identity. Not just to cheer on sports teams either. Individual faculties often have a cheer: Engineers have Godiva’s Hymn, and Math students apparently have “Hip Hip Array” and “Row, Row, Row, Reduce”.  Think of the unyielding songs and chants that we heard during FIFA World Cup this past summer, from national anthems to the unmistable “Ole Ole“. And to connect this to yesterday, singing is also a big part of collective labour, from the slave songs of American cotton fields to the office desks of IBM in the 1920 and 1930s.

Wait, what? Yes, that’s right, IBM, probably one of the most white-collar of American corporate employers in history, had an entire song-book for its workers to “express in happy songs the fine spirit of loyal cooperation and good fellowship”. That doesn’t sound like capitalism.

Nonetheless, the 54-page 1937 edition of “Songs of the IBM” contains numerous odes to the magnificence of IBM, its leaders, and of the great privilege and fortune that accrues from working for such a company. Here’s a taste:

“Ever Onward”
There’s a thrill in store for all,
For we’re about to toast
The corporation that we represent.
We’re here to cheer each pioneer
And also proudly boast
Of that “man of men,” our sterling president.
The name of T. J. Watson means a courage none can stem:
And we feel honored to be here to toast the “I. B. M.”
Thomas J. Watson was the first president of IBM and a powerful corporate titan. Did he need a cheering section? I don’t know.

As was pointed elsewhere online, IBM has made a few of the songs available online at http://www-03.ibm.com/ibm/history/exhibits/music/music_clips.html, which includes “Ever Onward” above.  And the Computer History Museum has a short video about IBM culture online at http://www.computerhistory.org/revolution/punched-cards/2/12/2211.

I can’t say that I’ve ever worked for a private company with it’s own cheer, but I wonder if this persists and what it could be like today. Maybe Facebook exhorts its workers to new invasions of privacy with a daily chorus. Do the assembly line workers in Apple factories in China begin each shift with solemn chant to commemorate and invoke the blessing and power of the deceased Apple icon and former CEO, Steve Jobs? And what about the hungry start-ups? You can’t code for 12 or 14 hours straight  without good rhythm, so instead of perks or stock-options, maybe what they should provide is a constant beating drum to drive the work forward!  Apps don’t code themselves.

Maybe CSTV needs its own “fighting” song. Any volunteers?

Fauxtography and computer graphics September 2, 2014

Posted by Cameron Shelley in : STV100 , comments closed

One thread we pick up in this blog from time to time is that of fauxtography, that is, faked photographs. Fauxtography is an interesting practice in its own right, and it also raises some philosophical issues about the authenticity of photographic images.

A couple of recent news items relate to this issue. First is a project described in The Atlantic of “de-specializing” Star Wars. As you may know, George Lucas has been in the habit of revising and re-releasing the first three movies of this series from time to time. Some fans, however, prefer (or just want the option to see) the original versions. So, they have undertaking a demanding and elaborate process of removing the alterations from later sources (Lucas has not allowed access to any original prints). This project is called “de-specialization”.

You can find out more about this project from this YouTube video by “Harmy”, one of the project’s leaders.

Note that many of of Lucas’s CG amendments have been removed not by restoration of an earlier version but by CG simulations of that version. So, the de-specialized version is really a simulation or facsimile of the original release. Is it authentic, or not?

It turns out that George Lucas is not the only fan of CG in photography. A recent item in FastCompany points out that about 75% of the images in Ikea’s catalog are “faked”. That is, they are CG. And that is up from just 25% two years ago. It turns out that making up images with computer graphics is just cheaper and easier than shipping products to photographers’ studios for physical shoots.

Another reason is that CG images are so good that they are hard to tell apart from the real thing. So, customers are seeing images that simulate real photos and are hardly distinguishable. Are they authentic?

In the Star Wars case, you could point out that the de-specialized versions are not genuine, that is, they are not copies of original prints of the films. However, they are (ideally) intended to be indistinguishable from genuine copies (or, more precisely, performances of genuine copies). Perhaps being indistinguishable from a genuine copy is enough for authenticity? After all, people accept “authentic” Greek yogurt that does not come from Greece. (Although, the EU holds that “champagne” must come from Champagne.)

In the Ikea case, there is no original image to simulate. Thus, the faked catalog photos are not genuine because they are not copies of original photos. Or, you might say that the CG images are “original” and thus the catalog photos derived from them are genuine after all. Perhaps the concept of genuine is not helpful. In that case, you could argue that the catalog photos are inauthentic because they appear to be something that they are not. Or, perhaps they are authentic because they are true-to-life.

Thoughts?

  • Regulus by Ben @ Binary Moon
  • Created with WordPress
  • Top
  • -->