Technological buzzwords like ‘Big Data’, the ‘Internet of Things’ and ‘Machine Learning’ are becoming ubiquitous in our increasingly techno-centric society, with advocates heralding the birth of a new data-driven age where predictive and analytic technologies will inform every aspect of our lives and lead to dramatic improvements in every sphere (be it healthcare, energy-efficiency, social care, waste management, or many others). Critics argue that we could soon be placing too much faith in the power of algorithms: in our power to write and develop the correct ones and in them acting benignly for our interests [1].

Due to the inundation of positive messaging about new technology in our culture, it has been very easy for society to become blindsided to the possible dangers lurking in our adoption of applications and software with easy access to data across our lives. Despite the benefits we may derive from this new technology, it is important to remember that like everything else, it does not exist in a vacuum and in many cases has been developed by powerful economic interests who seek ever-increasing profits, even if it is to the detriment of humanity.

Data is Power

An article on the ‘Jacobin’ website in 2015 [2] highlighted several areas where powerful groups are already using the latest wave of data-driven technologies in order to further their interests, irrespective of the effects on other sectors of global society.

For example, the article discusses how the use of statistical analyses by the U.S. military has already led to the deaths of several civilians at the hand of drone attacks [3]. Although the drone ‘Kill Cycle’ is required by International Law to be executed by humans, the analyses which lead to the targeting of individuals in these attacks has become increasingly abstract and computer driven. Jeremy Scahill, founder of the Intercept.org and and author of “Dirty Wars: The World is a Battlefield”, a critical account of the Obama administration’s use of drone warfare for killing suspected terrorists, discussed on Democracy Now how statistical analysis of SIM card usage metadata is often enough for the U.S. military to extra-judicially sentence suspected terrorists to death [2]. There are two concepts at play here which allow for such horrific consequences and must be challenged by wider society:

1. The belief by the U.S. military that the power of data is absolute, i.e statistical analysis is the best, and even only, way to arrive at the correct conclusion.

2. The fact that an entity such as the U.S. military has both unfettered access to such data in the first place and also that it does not have to answer to any legal or regulatory framework when drawing its conclusions i.e law (international and national) has not caught up with the technology.

Abuses of the power derived from data can of course also be found in the commercial sector, such as in the creation of ‘secret consumer scores’, which determine financial risk based on exceedingly simple data such as where a consumer lives [2] (this geographical discrimination will obviously intersect with class and race discrimination as well, thus further acting to entrench societal inequalities). The use of “scheduling software” by corporations also allows them to reduce working hours while at the same time increasing the amount of work performed by the employees, increasing profits but disadvantaging workers [2].

These are only some of the possible abuses of data-driven technologies, but they serve to highlight that talk of ‘big data’ as some sort of inherent positive is sorely misleading. For more information about this aspect of big data and its potential uses and abuses, I thoroughly recommend the Jacobin article referenced above.

Although it is important to challenge the ‘technological-messiah’ status assigned to data science, big data and associated technologies and capabilities, and to highlight how they are currently being used to further the interests of military and corporate organisations, this is not the only thing progressives must do. It is important to recognise that the abuse of such technology is not inevitable, it can and should be challenged.

I will now go on to discuss the possible emancipatory power of these new technologies and the associated cultural and societal shifts that progressives should aim to bring about in order to maximise their benefit for humanity.

Robots Will Take Your Job, and You Should Let Them

Thanks to increasing coverage in the media, most of society is now aware of the potential for large-scale technological unemployment [4-6], whereby the growth of artificial intelligence, machine learning and robotics will lead to the replacement of workers with automated systems. A working paper from the Martin School at the University of Oxford on “The Future of Employment”, suggests that approximately 47% of US jobs are at risk of automation in the near future [7]. Other publications have put this figure as high as 80% of US jobs [8]. Clearly the potential for technological unemployment is huge and pressing.

In their book “Inventing the Future: Postcapitalism and a World Without Work” [9], Nick Srnicek and Alex Williams highlight what they see as the likely outcomes of an increase in the level of technological unemployment brought about by an increase in automation through robots and data-driven software:

“1) The precarity of the developed economies’ working class will intensify due to the surplus global labour supply (resulting from both globalisation and automation).

2) Jobless recoveries will continue to deepen and lengthen, predominantly affecting those whose jobs can be automated at the time.

3) Slum populations will continue to grow due to the automation of low-skilled service work, and will be exacerbated by premature deindustrialisation.

4) Urban marginality in the developed economies will grow in size as low-skilled, low-wage jobs are automated.

5) The transformation of higher education into job training will be hastened in a desperate attempt to increase the supply of high-skilled workers.

6) Growth will remain slow and make the expansion of replacement jobs unlikely.

7) The changes to workfare, immigration controls and mass incarceration will deepen as those without jobs are increasingly subjected to coercive controls and survival economies.”

In the developed world, many of these outcomes are already manifesting themselves, especially: the increased precarity of work; the transformation of higher education institutions to become solely job-training centres; and the increase in incarceration of the poor. This paints a very grim picture of the future, but it is important to realise that these outcomes are not inevitable. In fact, the same trends in technological automation can be harnessed for progressive causes and for bettering the lot of people across the globe.

This is discussed in the rest of the book when Srnicek and Williams go on to argue that progressives of all stripes can seize this moment. While technology is increasingly acting to disrupt societal norms, it can be used also to challenge, disrupt and replace the current hegemony of neoliberalism. In order to make this a reality they argue that progressives must demand certain concessions from government and push for a change in  society’s attitude to work. These can be summarised thus:

– Demand and push for a fully automated economy.

– Reduce the length of the working week with no cut in pay.

– Demand the introduction of a Universal Basic Income that supplements, not replaces, the welfare state.

– Move towards the diminishment of the work ethic, which permeates society and treats those to do not work as ‘feckless’, ‘lazy’ or at its worst, sub-human.

The authors argue that these deceptively simple demands push at the limits of what is conceivable within the current window of allowed political thought and will act to accelerate the replacement of our current economic and social system with a new, far more progressive one. It is important to stress here that Srnicek and Williams’ thesis is not one which supports the current trends of automation. As it stands, automation combined with a huge global labour supply, means that companies can increasingly decrease the number of jobs available, thus encouraging intense competition for work and pushing down wages substantially (zero hours contracts are a prime example). The corporations which own the technology are therefore making the labour market ever more ‘slack’, benefitting their bottom line substantially. Srnicek and Williams argue that if we can enact the four demands above, then the power dynamics between labour and capital fundamentally change. The universal basic income (at a high level and on top of current benefits) will mean people need not compete so intensely for what jobs there are, the sharing of work through the diminishment of the working week acting likewise. This therefore means that with increased automation the labour market actually becomes ‘tighter’, so there is more jobs going than people competing for them (as most will not need to work) therefore increasing labour’s bargaining power against capital and driving up wages for what few jobs there are. The diminishment of the work ethic further enhances this trend such that as we reach full automation (really 47-80% of all jobs as stated above) the power dynamics have been completely turned on their head and those with capital no longer have the power to wantonly exploit working people. This is automation, but not as we know it currently.

On the whole I agree with Srnicek and Williams’ analyses and believe that their proposed demands are both achievable and desirable, and I would encourage readers to read their book. Automation is already happening, so pushing to accelerate this trend is very achievable. Governments are already considering implementing UBI’s, so applying pressure to ensure this is done in a way which simply does not replace the welfare system is possible. A cut in the working week without a cut in pay can very likely be touted as a cross-political spectrum stop-gap to distribute and save what employment opportunities we can whilst machines increasingly take over the labour market. And finally, the diminishment of the work ethic, although difficult in a society riddled with right-wing demonisation of the (working and non-working) poor, may become increasingly likely as more and more people across traditional class divides find themselves out of work due to automation.

The problem comes in concretely defining what fully replaces neoliberal capitalism, rather than simply highlighting the fronts to push on to permanently weaken it (t­­­he four points above). Srnicek and Williams, like Paul Mason before them in “Postcapitalism: A Guide to Our Future” [9], fall into the trap of promulgating heavy critiques of now while not offering a fully formed view of the future they want to see. To give credit where credit is due, Srnicek and Williams do a better job than other literature I have read in identifying the next steps progressives should take in order to create a ‘brave new world’, even if they do not give a view of the ultimate destination. To tackle this point, in a future article I will discuss the possibilities of what a “postcapitalist” economic system could look like. For now, however, I have hopefully made the case that the four points expounded by Srnicek and Williams above, when pursued in tandem, can constitute a progressive response to increased automation.

What Can You Do Now? Using Data and Science for Good in the Near Term

This article has highlighted the tendency of powerful interests to abuse data and new technology, as well as proposed political and cultural responses which can harness them for good in the medium to long term. But other than discussing the longer-term politics of data and technology, what can scientists and data savvy individuals do to help progressive causes and organisations right now? We at PSI have been researching this question and would like to suggest the following possibilities (we are of course open to more, so if you have any ideas please get in touch):

1. There are several organisations which could benefit from relatively simple analyses of publicly available data. For example “Bath Hacked”, a local group of data and machine learning enthusiasts in Bath, have been studying data on bicycle rides available from Strava Metro [10,11]. They have produced an interactive dashboard which allows users to interrogate the results, for example by highlighting junctions which cause delays for cyclists [11]. We are very keen to study similar data for cities across the UK – there is currently data available from Strava Metro for London and Glasgow, which PSI should be able to access as an advocacy group, in order to approach local government with proposals for improved cycling infrastructure. If you are interested in helping with this project then get in touch. (We are also considering the possibility of providing data-driven insights which may aid an increase in recycling as well as encouraging the development of more environmentally friendly city infrastructure).

2. The technology driven non-profit NGO, “Bayes Impact” [12], have highlighted several cases on their website where they say they have developed data-driven technology in order to improve public services across the world. Although we at PSI have not studied their work in detail, we believe there are many exciting possibilities in applying the same ideas to public sector challenges here in the UK (and further afield).For example, data science has already been utilised by charities such as the Trussell Trust to optimise provision of resources from foodbanks [13]. This, however, is unsatisfactory since it is treating a symptom of inequality and not the cause, we are therefore keen to explore the possibility of providing data driven analysis of public sector and social care data in order to help local organisations and charities attack the cause of such issues. Local governments are already waking up to the possibilities of using data skilled people in this way (see this article in the Guardian, which argues for more funding for councils to hire data scientists [14]). We at PSI would like to find ways to support similar initiatives and use data skills to tackle social issues. If you are interested in helping with this project then get in touch.

3. We would also like to encourage those with data skills to apply these to citizen science projects in order to aid the analysis of climate change data and provide support for those organisations which are aiming to lobby policy makers (such as Greenpeace UK and Friends of The Earth). Studies of solar power and wind potential across UK cities could provide an excellent starting point for approaching local government and asking for support for local renewable energy projects. We believe a more local approach like this will be vital to counter the UK government’s antagonistic stance toward renewable energy. Again, if you are interested in helping with a project like this then get in touch.

References

[1] “Digital Star Chamber – Judge, Jury and Executioner, the unaccountable algorithm”, Aeon, 18/8/2015, https://aeon.co/essays/judge-jury-and-executioner-the-unaccountable-algorithm

[2] ‘Big Data’s Radical Potential’, Jacobin, 3/12/2015, https://www.jacobinmag.com/2015/03/big-data-drones-privacy-workers/

[3] “Turning a Wedding Into A Funeral: U.S Drone Strike in Yemen Killed as Many as 12 Civilians”, Democracy Now, 21/2/2014, https://www.democracynow.org/2014/2/21/turning_a_wedding_into_a_funeral

[4] “Will a robot take your job?”, BBC News, 11/9/2015, http://www.bbc.co.uk/news/technology-34066941.

[5] “Elon Musk: Robots will take your jobs, and government will have to pay your wage”, CNBC, 4/11/2016, http://www.cnbc.com/2016/11/04/elon-musk-robots-will-take-your-jobs-government-will-have-to-pay-your-wage.html.

[6] “Would you bet against sex robots? AI ‘could leave half of world unemployed’”, Guardian, 13/2/2016, https://www.theguardian.com/technology/2016/feb/13/artificial-intelligence-ai-unemployment-jobs-moshe-vardi .

[7] “The Future of Employment” , pg 38, Frey and Obsorne, Sep 2013, http://www.oxfordmartin.ox.ac.uk/publications/view/1314.

[8] ‘Anticipating a Luddite Revival, Stuart Elliot, Issues in Science and Technology, 30: 3 (2014), http://issues.org/30-3/stuart/ .

[9] “Inventing the Future: Postcapitalism and a World Without Work”, Nick Srnicek and Alex Williams, Verso Books, 2016.

[10] “Postcapitalism: A Guide to Our Future”, Paul Mason, Jul 2015, Penguin.

[11] Bath Hacked, https://cyclebath.org.uk/2016/03/25/is-there-value-in-the-bath-strava-data/, Bath Hacked study Strava Metro data, https://cyclebath.org.uk/2016/03/25/is-there-value-in-the-bath-strava-data/. Their dashboard can be found here http://strava.bathhacked.org/interact/#13.67/51.3823/-2.3683 .

[12] Strava Metro, http://metro.strava.com/.

[13] Bayes Impact, an NGO and non-profit who use Big Data to solve problems in Health, Unemployment and Justice through building data driven social services. http://www.bayesimpact.org/

[14] “How data science is helping charities to fight hunger in the UK”, 9/5/2016, Guardian, https://www.theguardian.com/voluntary-sector-network/2016/may/09/data-science-helping-charities-fight-hunger

[15] “New Year, new career: how about becoming a local government data scientist?” , Guardian, 10/1/2017, https://www.theguardian.com/public-leaders-network/2017/jan/10/new-year-career-local-government-data-scientist

Data is Not Neutral – How Powerful Interests are Shaping 21st Century Technology and Robots, Not Immigrants, are Taking your Job
Tagged on:             

Leave a Reply

Your email address will not be published. Required fields are marked *