Thread by @nettwerkerin

Reading recommendation for @zararah 's piece on data and identity: "Can data ever know who we really are"?
deepdives.in/can-data-ever-…

Find below some quotes and comments on this thoughtful article.

The article starts with the description of an intro question that all of us have faced and her reaction to it:

Contrast this with the quote by Mark Zuckerberg, who said (according David Kirkpatrick in his book "The Facebook effect"):

"Having two identities for yourself is a lack of integrity".

@zararah stresses that identity is context-dependent and dynamically changing - while almost all digital profiles are rather static and context-independent.

She gives the example of the Fitbit and its limited menu choices when creating the user's profile, e.g., a binary gender option. The choices will reflect some of the results that the Fitbit makes out of the sensory data (in ways that are rather intransparent to the user).

She quotes her mother: "‘I slept badly,’ my mother says. ‘My Fitbit says so.’" and makes the point that we believe more in machines and their seemingly precise data than in our own feelings. She sees the cause for this in something called 'precision bias'.

I am currently reading Kees van Deemter's book on "Vagueness" (which I also recommend) and found something interesting there which might explain this "bias". As a computer linguist, he mentions the "maxim of quantity" in his book "Not exactly - In praise of vagueness", p. 116:

"Express yourself as strongly as your information allows.". He calls it: "one of the most important (principles) to come out of modern linguistics so far".

If we all humans follow that rule, it is the machine that is miscommunicating, and not us being biased.

@zararah stresses that the belief of her mother's in the result of her Fitbit is especially weird given the limited choices for a limited number of categories to describe yourself.

Here is one place where the mindset of software designers can enter the software: by making a choice of the categories by which users can describe themselves. What they do not know or not deem an important category, is lost. Rahman writes:

@zararah summarizes that there are hard limits to categorization and thus, to digitalization as well. The article ends with the sentence:

"Sometimes, complexity is a feature, not a bug."

@zararah @threadreaderapp please unroll