This is not your great-grandfather’s industrial revolution
Fri 27 May 2016
Editorial We had a quite extraordinary response this week to an article outlining Foxconn’s replacement of 60,000 jobs with robots. What society will do in the wake of automation and AI is currently among the hottest tech topics, and as I write, the former CEO of MacDonald’s – the poster-child for mass consumer economics – has weighed in further on the issue by auguring the transfer of the McJob to, quite literally, a robotics arm.
Many have observed a recursive theoretical problem with the automation of large-scale industry and consumer services: the underlying economics are viable because a large swathe of society can afford the products being produced, whether by people or machines. When the aggregate mass can no longer afford the aggregate product, the model theoretically breaks down.
The universal ‘citizen’s wage’ which is occasionally proposed seems unlikely to keep that model aloft, at least at the same economy-of-scale.
The answer in China seems to be the casual sop which the west has so often applied – that the displaced manual workers will move to service industries; the problem is that service industries are also seeking automated solutions that employ AI and robotics.
Could it be that the biggest commercial concerns on the planet are blindly (and suicidally) applying 21st-century solutions to 19th-century economic models in a short-term attempt to amass wealth before the model itself collapses under the innovation?
But the topic of automation is running along a separate and much older path than the neocon economics which have characterised the property-driven ‘recovery’ since 2008.
Controversies around automation don’t bear the same Dickensian hallmark as the latest furore over the gentrification of major cities such as London (which are currently sending economic refugees to outlying borders in order to sustain the capital value which has single-handedly underpinned the tentative recovery since 2008).
The hyper-valuation of city property treats ‘displaced’ natives as collateral damage in the construction of an increasingly exclusive club. But proliferating automation promises to undermine consumer society in a manner far more self-destructive than the latest economic short-cut – or the steam-driven jennies of the early industrial revolution.
Perhaps we should think about where that road might actually, finally run out.
Ron Nash of Pivot3 wrote an excellent piece for us last Monday on the rise of hyperconvergence in the data centre, and it’s a recommended read for those pondering what the next stages might be in the progression towards the software-defined data centre (SDDC).
Controlling off-the-shelf hardware architecture via a software layer, complete with scalability, predictive monitoring and analytics, is a significant step towards the default hardware agnosticism envisioned as standard industry practice in an SDDC-centric world; but it’s going to involve a major change of approach from the silo build to genuinely open standards. It’s the kind of quantum shift in planning and construction practices that’s going to need vanguard, headline companies to push forward into the mainstream.
As Ron points out, the financially prudent retention of legacy hardware is likely to prove one of the key obstacles to hyperconvergence and, by implication, SDDC. An organisation that’s institutionally accustomed to continually leveraging once-expensive, dedicated hardware with peripheral hardware and software upgrades is likely to wait upon wider adoption – an old story.
Elsewhere in the site we take a look at the worth of network monitoring in the data centre – a potentially expensive proposition for which the ROI case isn’t always clear, and for which business reasoning needs to be unusually clear-headed and forward-looking.
On a less ponderous note, Google has taken a leading approach to the potential drabness of data centre exteriors by commissioning large scale murals at several of its DC locations around the world. As Joe Kava, the vice president of Google Data Centers observes, “Billions of requests are sent to ‘the cloud’ every day. But few people know all this information flows through physical locations, called data centers. Because these buildings typically aren’t much to look at, people usually don’t, and rarely learn about the incredible structures and people who make so much of modern life possible.”