algorithm

What is Algorithm? Tarleton shown us the definitions of Algorithm in four different aspects:
1.
Algorithm as a technical solution to a technical problem - a logical series of steps for organising data to achieve a desired outcome quickly.
2.
Algorithm as synecdoche - in a more sociological angle, it’s a kind of socio-technical ensemble, part of a family of authoritative systems for knowledge production or decision-making.
3.
Algorithm as talisman - it’s like a trick for information industries, to persuade the audience that the product, or saying the designer/corporate owner makes objective, legitimate decisions.
4.
Algorithmic as committed to a procedure - In general, everything that is driven by and committed to algorithmic systems is algorithm.

To be honest, this reading made me feel like everything in our society could be algorithmic: products we use daily like Facebook, Spotify, Tinder, Instagram etc. Everything related to statistical data is algorithmic. And we can’t really run away from those “number facts”, even when we made decisions by ourselves, we consider the social norm that generated ethnographically, as well as the one generated by algorithm.

When something is algorithmic, it’s a formalization of social facts: make them into “measurable data”, and make the social phenomena “clarifiable”. So there would be always problem and solution. Like certain conditions leads to certain result, like the “if/then” logic in the technically defined algorithm.

But is everything measurable? Does every problem have its corresponding solution? When we think about something organic, something emotional. Algorithm seems a bit stark and inhuman. Like nowadays some people would like to see their body (organ) data to estimate their health situation. When we say ‘you’re out of shape because your body fat percentage is too high’, it still sounds reasonable. But when we say ’you’re angry because your epinephrine/heart rate number is raising too fast’, it sounds not communicating anything at all. (‘You’re angry because your friend didn’t answer your phone’ sounds a bit more communicable.) So seems generally, algorithm is not very useful for something organic and emotional. (However, if via algorithm, we applied those organ data to generate visual/audio/something people can sense with, it’d actually function well in these emotional communications.)

To prevent people from falling into a confusion that everything is algorithmic, Tarleton also talked about what is not algorithmic:
- something done subjectively,
- something done by hand,
- something can only be accomplished with persistent human oversight,
- something is limited by context, etc.

So I started to think about the examples. The first few things I came up with is photography, painting, telling stories. Traditionally, these are very subjective and creative behavior. They are more about personal experience, personal aesthetic, intuition, and perception, rather than logic and statistical numbers.

However, another extremely contradictory example crossed my mind: Botnik, a community of writers, artists and developers using machines to create things, fed the seven Harry Potter novels through their predictive text keyboard, then it came up with a chapter from a new Harry Potter story: Harry Potter and the Portrait of What Looked Like a Large Pile of Ash

Then following this clue, I found another AI writing example: Director Oscar Sharp and AI researcher Ross Goodwin used a recurrent neural network called long short-term memory, or LSTM for short to write a film script, and made the film Sunspring, for Sci-Fi London, an annual film festival that includes the 48-Hour Film Challenge. The process was quite similar with Botnik’s Harry Potter. They fed the LSTM thousands of sci-fi film scripts, some random seeds from a sci-fi filmmaking contest including a title (sunspring), a piece of dialogue, a prop and an action, and an optional science idea, then the AI started writing the play.

Although such ‘AI writing’ are just reflecting on and simulating the human-written texts we’ve fed them, there’s actually nothing ‘creative’. But I have to admit that I found some of those contents are unpredictable interesting, or even ineffably poetic to me. So I guess the random composition created by the algorithmic could be organic, emotional as well (depends on different people’s different interpretations).

In summary, as Tarleton said in his Algorithm [draft], the advantage of algorithm is its democracy, systemisation, and we also concern that it might be an extension of Taylorism, it might be too stark and inhuman. But from some computational art works/ practices, I saw the ineffable human side of algorithm, and I think that’s what we are working for, breaking the boundaries, collaborating with the machine, and showing better humanity.

Reading materials:

Algorithm [draft] [#digitalkeywords] by Tarleton Gillespie
Harry Potter: Written by Artificial Intelligence
Movie written by algorithm turns out to be hilarious and intense by Annalee Newitz
Illustrations from Megan Nicole Dong