Category Archives: Blog Posts By Me

For EDC students who may wander on to my blog. These are the posts I think would be most suited for others commenting. Of course, you are welcome to comment on my other posts if you want.

Neat definition of Cyberculture

Frow and Morris (2000: 316), define culture neatly as ‘a network of embedded practices and representations (texts, images, talk, codes of behavior, and the narrative structures organizing these) that shapes every aspect of social life’. Cyberculture therefore refers here to ways of life in cyberspace, or ways of life shaped by cyberspace, where cyberspace is a matrix of embedded practices and representations.


Lifestreaming and mental health

So from the about lifeblogging link suggested in the course handbook I ended up looking at the Quantified Self conference, this guy presented his deeply wrongheaded and creepy software at the conference:

From the off, I oppose the ultimate assumption he makes that life is about maximising your happiness. This is overly utilitarian and needlessly devalues the worth of other emotional experiences that are essential to the human experience. Putting the futility of his overall aim aside, I have three other criticisms, to whit;

First, it is worth questioning whether happiness can be measured in the ways he suggests. At several points in his presentation he mentions in passing that the measurements can be adjusted to the individual depending on the weighting they give each metric. So the usefulness of his software relies on the mindboggling assumption that a person can know exactly what makes them happy and by how much in terms of opportunity cost.

Second, the measurements totally ignore personal context. For example, travel is assumed to increase your mental well-being but what if you have to travel across the country to attend the funeral of a relative? This would increase your score but would it really make you happier?

Third, the measurements also ignore societal context. By relentlessly focusing on personal data Milburn ignores the possibility that humans are social animals and that our mental health can be affected by our position in a wider society. Status anxiety exacerbated by societal income inequality can be a major factor in mental health (, yet this is not reflected in his metrics.

What this boils down to is that Milburn confuses correlation with causation and then solely focuses on the factors that tracking technology makes the easiest to quantify.

As ever, this leads me back to music. A soundtrack:

“Karma Police arrest this man, he talks in maths, he buzzes like a fridge, he’s like a detuned radio”


Possible implications of google’s improved translation AI:

1.) Less paid work for professional translators as there is now a free service that can provide near human translation


Q: Will google translate ever be accepted as an “official” translation by governments?

When my wife applied for her visa we paid a considerable sum to have an official, government approved translation of our flat lease in China (even though my wife could have done just a good a job translating it). Technologies fit with current social structures is an important factor in whether they are adopted on mass.


2.) Potential monopolies and force feeding the algorithm.


As it stands google does not charge for its’ translation services. If the AI is really that good that it gradually replaces other translation services do we not risk giving too much market power to google. What’s to stop them from charging once the market place is cleared out?

Also, although the service is free what google gets from you is more data for the algorithm to work from. Presumably google then owns the original text you inputted as well as translation and will do what they want with it. This is quite different from professional, human based translation services where there are ethical guidelines preventing the retention of client’s documents. Is that a price we are willing to pay?



3.) Easier translation but loss of nuance and cultural knowledge? Does this matter as long as it is “good enough” for quick jobs?


Just because the words are translated accurately into your language doesn’t mean that text can be understood. Metaphor, similes, cultural references, all of these require extra knowledge which a free translation service cannot provide. For example, you could translate 知音 as “bosom friend” as google translate currently does.  This is a perfectly functional as a translation, within one context. But 知音 can also be translated as “soulmate” and has an elaborate etymology that google translate does nothing to reveal.

知音 refers to story from the warring states period. A nobleman called Boya was playing his guqin when he was overheard by a simple woodcutter, Zhong Ziqi. To the nobleman’s amazement this common and unsophisticated woodsman could appreciate his refined playing and describe exactly what images were in the nobleman’s mind when he played. The woodcutter and the nobleman were each others’ 知音. The person who knows the sound of your heart. A different English idiom could apply here “someone who speaks my language”. But you don’t get that option from google (it was from teaching my wife this idiom that led her to teach me 知音). I know enough Chinese to know to do a web search for 知音 and remind myself of the etymology but someone using google translate who hadn’t spent time trying to learn Chinese wouldn’t know to do that.

Doing a google web search for 知音 led me to some interesting results. I’ve given you my version of the 知音story above, here is someone else’s version, here is another. They all tell the same story but in different ways and with different nuances. The third story simplifies the implied social relationship between the Boya and Zhong Ziqi to an overly blunt “they had nothing in common”. Now I also know that guqin is an instrument that is strongly linked with upper class literati and so I tried to get across the class difference between the two characters in a subtler fashion.

As this story dates back to the warring states period there is no definitive version, all three interpretations are valid. However, this kind of complex “fuzzy” epistemology is anathema to algorithmic thinking. In order to be incorporated into an algorithm meaning has to be made precise so that the data can be tagged and searchable. This can be problematic, not only in terms of epistemology but also with regards to hidden agendas for excluding data. For instance, let’s assume that google translate is going to become even more sophisticated and can eventually start to draw upon folklore and etymology to give flexible, nuanced translations. Which version of the 知音 story does it choose as the “correct” story for its database? As google has a financial motivation to increase revenue from targeted advertising they would want to keep you within its network of services, thus it is feasible they would select the third story for their database as it is published on googleplus. This goes against my judgement of the third story as the least helpful for understanding 知音. Relevance is entirely subjective and problematic for algorithms.