Data is not Memory
“Data is Data” explored that data is not oil but that data is something that is unique and we should call it for what it is; #Data. The reason explained is that all the analogies: oil, gold, sunshine, commodities, time and labour, fail to bring out the uniqueness of data. Whilst the thinking explored in that post still stands, a year on I have to challenge one key assumption, which I will come to.
I recently posted ideas of where memory comes from, and specifically the links between memory and identity. We have different types of “memory” for myself, society and nature and now data/ digital might be creating a new type of memory. The reason memory and its link to data is worth exploring is that both depend on relationships to derive “value.” Data without relationships is worthless. Memory without relationship (connection) is worthless.
The assumption that we need to look, “In our complex digital world what type of “memory” do we want the system to have?” Memory in this case being the collection and storage of everything forever, with perfect recall, and the system being those who create/extract value from the collection, storage and analysis of this data.
We assume/ understand that nature does not have a permanency of memory (no perfect short or long term memory) and therefore citizens tend to assume/ think that all systems must have the same loss and erosion of the past (data). However the digital data collectors are not following nature in this regard; insomuch that they collect and store everything in perfect techi-colour. The justification/ view being that the more data we have, the better I can know you better than you know yourself. Ultimate personalisation from perfect data and prediction.
I have explored many times that this is NOT a view I agree with now (I did in 2009); more data does not make you more or completely predictable. Data certainly identifies patterns and behaviours and these are repeatable, but that does not make you predictable. We also dont collect all data and we don’t currently have a model for the prediction, however, in the process we are creating mass data stores of our digital twin and the question is surely what future will this collection of data bring about?
Since the very earliest times of writing and philosophy we have tried to overcome death and determine end games. Back in early writings there was no complex end game scenarios, it was very simple and often singular in outcome by culture or geography. In more recent times we have been able to deal with more and more complexity leading to more possible end game scenarios. Whilst the dystopia views from 1984 to Logan Run provide insights into surveillance and control; the “Minority Report” thinking was more subtle as it explored using data to prevent harm, but the key assumptions are:
- perfect data can be collected
- perfect data creates perfect prediction
- the perfect prediction can be prevented - changing the system without affect
The observation is that this is one outcome from the ideals of perfect data, however we have not explored all the possible and probable outcomes and surly in 2020 it is something we should do. We have to unpack outcomes of perfect data as this is not nature, and therefore we cannot assume that nature will give guidance to the outcome. Perfect data will not drive out human ills ( 7 deadly sins) but what will perfect data give us to create a better society?