• Instagram

©2020 amiclarke

Low Animal Spirits by Ami Clarke and Richard Cochrane 2014.
Installation at Banner Repeater - with support from The Elephant Trust and ACE.

 

It employs a high-frequency trading (HFT) algorithm written by Cochrane who was once a Vice President of Goldman Sachs, and ‘deals’ in words sourced from global news feeds for virtual ‘profit’, whilst speculating on their usage. It is driven by real-time data, and scripted as a live onscreen score and audio work with automated readers. The analysis produces new phenomena in the form of speculative headlines tweeted from the twitterbot @LowAnimalSpirit. The projected visualisation is a glimpse into the HFT algorithms buying and selling activity with volatility at the centre of these operations. It accesses 994 English Language global news feeds ‘live’ and acts upon the data as if it were trading in the global market place, analysing words in terms of the potential for a virtual ‘profit’ to be made. The resulting visualisation displays what is about to trend, and likewise, the speculative headline generator tries to anticipate the next headline, based on recent history, and incoming headlines, with the help of a Natural Language Processing algorithm, tweeting from @LowAnimalSpirit every two minutes. The work produced some unexpected results that were made visible and produced through the specific visualisation process that we had chosen. During periods of low activity, the HFT algorithm ‘sells off’ much of it’s portfolio, and then ‘buys’ many new, but small items, as a kind of sifting process. This behaviour change in the algorithm can be seen when many words briefly appear to ‘swarm’.  It was noted on several occasions that the work is a topology of sorts.


Low Animal Spirits was a phrase coined by the economist John Maynard Keynes during the Great Depression of 1929 and almost a hundred years later had cause to be cited1 again, notably by the governor of the Bank of England Mervyn King in 2012 (King and Flanders, interview 2012). Keynes had spent many years studying probability theory in an attempt to predict future market behaviour and the Great Depression shook his confidence considerably unsettling all his investments, theoretical and otherwise, in ideas of a self-regulating and sustainable market. ‘Low animal spirits’ – a mass downturn in confidence – showed a paradox found in human nature that locates a herd-like mentality, as well as unpredictability and uncertainty, at the core of international finance: a model of mass behavioural procedures. The complexities that arose, of the lone individual acting spontaneously within a crowd, conceived through statistical analysis and probability theory, serves to address complex conceptualizations of the individual within new formulations of the crowd. Since the phrase was first coined, the effects of probability theory have increased to an unprecedented scale, and the software/hardware, and infrastructural developments that facilitate these operations, plus political investment in de-regulation and a poorly legislated market place, have developed to the point at which human cognitive powers can no longer keep up.  The arena in which these ideas arose is of a time where the very concept of free will came into being during the Enlightenment and as Hayles (1999) notes; an evolving subjectivity that emerged through market relations. Gambling and games of chance, with all the romantic heroism of the autonomous subject pitted against mathematical determinism, drove the quest for ever better odds, whilst the eighteenth century study of celestial mechanics developed complex new mathematical tools for deriving probable outcomes. The term animal spirits was drawn from the Latin spiritus animals to describe the elusive energy, or spirit, at one point thought of as a fluid, that drives human beings and has associations with Hume’s concept of spontaneous motivation.


The focus on news – a social production interwoven with a history of technological innovation, from the era of the printing press to present day hybrid strategies that include both print and digital platforms/distribution – has an underlying requirement with the need for language to be tethered to an instance based on fact and consequently a focus on how truth may be defined.

Breaking News - Crash Flash.

@LowAnimalSpirit twitter bot - tweeting live speculative headlines

The analysis produces new phenomena in the form of headlines generated with the help of Natural Language processing algorithms, tweeted @LowAnimalSpirit.