Blog All Dog-eared Pages: Radical Technologies by Adam Greenfield

Another from the overlooked-drafts pile. Seems like this just stalled on me writing the intro blurb.

Radical Technologies by Adam Greenfield (on OpenLibrary) is (was? It’s probably missing some of the “AI” bullshit, and the cryptocurrency section thankfully is less important a few years on) an excellent primer on a whole range of recent waves of the technology hype cycle; all viewed through a suitably critical lens.

Page 23

[…] you’ll realize with a start that what manufacturers are generally pleased to describe as “intuitive” is in fact anything but.

Page 24

There are two aspects of [Google’s decisions over what to diplay (and what not to) on its maps] to take note of: the seamless, all-but-unremarked-upon splicing of revenue-generating processes into ordinary behaviour, which is a pattern that will crop up time and again in the pages to come, and the fact that by tailoring its depiction of the environment to their behaviour, the smartphone presents each individual user with a different map.

Page 89

Significant areas of the economy might stand to be reclaimed for the commons. [Sounds like a good plan to me.]

Page 90

But straightforwardly, making things close to where they’re needed opens up the possibility of a denser, more compact and efficient way of living in cities. And with clean, city-center worksops sited cheek-by-jowl with living quarters, even urban planning’s basic distinction between industrial, commercial and residential zones comes into question.

Page 117

For all the hype around Bitcoin, it is clear that in its design, important questions about human interaction, collaboration and conviviality are being legislated at the level of technological infrastructure. its appearance in the world economy gives disproportionately great power to those individuals and institutions that understand how it does what it does, and are best able to operationalize that understanding.

[I don’t think it occurred to me at the time, but that comment about power could equally apply to the Internet. The main difference seems to be the lack of people trying to spread that understanding and power.]

Page 172

[…] the smart contracts on which DAOs are built, by their very nature, render decisions in the present on situations that were conceptualized at some arbitrary point in the past. In other words, a smart contract intervenes in a state of affairs that may have evolved in ways that were not forseen by the parties to it at the time they agreed to be bound by its terms, and does so irreversibly.

Page 194

As far as industry is concerned, though—and in this instance it really is their perspective that weighs heaviest and counts most—automation also means far less elaborate technologies, like the touchscreen ordering kiosks McDonald’s began introducing into its locations in the fall of 2014. In fact, automation means anything that reduces the need for human workers, whether it’s a picking-and-packing robot, a wearable biometric monitor, a mobile-phone app or the redesign of a business process.

Page 195

This shrunken workforce will be asked to do more, for lower wages, at a yet higher pace. Amazon is again the leading indicator here. Its warehouse workers are hired on fixed, short-term contracts, through a deniable outsourcing agency, and precluded from raises, benefits, opportunities for advancement or the meaningful prospect of permananent employment.

Page 196

Most of the blue-collar workers that do manage to retain employment will find themselves “below the API”—that is, subject to having their shifts scheduled by optimization algorithm, on little or no notice, for periods potentially incommensurate with their needs for sleep and restoration, their family life, or their other obligations.

Page 199

The prejudicial findings of such “HR analytics”, i.e. that a given employee is unreliable, costly or a litigation risk, may be acted upon even if the algorithm that produced them is garbage and the data little better than noise.

Page 205

On our way to a world of total automation, we may often have time to contemplate what a society winds up looking like when its most mutinous voices have fallen silent.

Page 206

All too often work cost us our health, our dreams, our lives. But it also offered us a context in which we might organize our skills and talents, it gave us some measure of common cause with others who laboured under similar conditions, across all bounds of space and time, and if nothing else it filled the hours of our days on Earth. Though these goods came at far too high a price, I don’t know that we are wise to consider living entirely without them, or are practically prepared to do so.

Page 226

[…] the more people affected by a particular act of automation, the more vulnerable those people are, and the harder it would be to reverse its effects, the more cautious we should be in enacting it. Our task as a society would then be to determine just where in this envelope any given proposed displacement lies. By these lights, we ought to have a great deal of concern when someone is proposing to bring learning algorithms to bear directly to bear on decisions of great public consequence, on a population that is already at risk, with immediate and life-changing consequences.

Page 238

You can teach an algorithm to recognize a table readily enough, based on its characteristics and the ways in which it relates to the world’s other contents. it might be able to identify, with successively finer degress of precision, a vehicle, a car, a police car, a New York City police car. That’s straightforward enough. But how do you teach it to recognize poverty?

Page 244

Quite simply, some parties derive advantage from the fact that we don’t understand the tools used to rank and order us.

Page 256

What comes to be the object of belief, in short, resculpts the space of possibilities we’re presented with. The conviction that autonomous operation isn’t merely possible in principle, but actually imminently practicable, operates at multiple levels, and creates multiple kinds of consequences. I think it’s by now reasonably well understood that the truly vexatious complications of automation are almost never technical but legal, regulatory, institutional, and those invariably take longer to settle out than any mere matter of invention and development.

Page 280

Whether intended or otherwise, one of the primary effects of the Stacks’ investment in young and emerging technical talent is to create a robust market for high-risk innovation with equally high “upside potential.” At any given moment, there are thousands of startups busily exploring the edges of technological possibility, and shouldering all of the risk involved in doing so. If their ideas come to nothing, so do they; they fade from the world without any further ado, and perhaps disperse their talent to other ventures. Should one of those fledgling concerns come up with a technique, a process or a useful bit of intellectual property, however, they will wind up being courted by one or more of the Stacks, with an eye toward eventual consummation in purchase. In fact the technology doesn’t have to be anything flashy, so long as it shows insight or promise; the Stacks routinely acquire startups not so much because they need access to a particular technique, but because strategically denying their competitors design talent is a cost-effective way of preempting them.

Page 284

For all the weirdness and vitality percolating up from the bottom of the technological food chain, a profoundly conservative tendency reigns at its apex.

Page 299

As individuals and as societies, we desperately need to acquire a more sophisticated understanding of how technologies work in the world, and who benefits the most from the way they accomplish that work. In part, this means applying the tools of institutional and discourse analysis to the technical innovation ecosystem, both to prise out latent patterns of interest and to demonstrate that certain statements and framings are the product of interest in the first place. But part of it is just learning to ask the right questions whenever we’re presented with a new technological proposition.

Page 302

This is what the great British cyberneticist Stafford Beer meant when he argued that “the purpose of a system is what it does.” On this theory, it doesn’t matter whether some technology was intended by its designer to enslave or to liberate, to preserve or to destroy. All that matters is what it is observed to do, and we ought to evaluate it on that basis alone.

[…]

And if given technologies cannot be evaluated at the level of their designers’ intention, we need to be still more wary of the promises made to us by developers, promoters and others with a material interest in seeing them spread.

Page 304

In every case the hard, unglamorous, thankless work of building institutions and organizing communitites will demand enormous investments of time and effort, and is by no means guaranteed to end in success. But it is far less likely to be subverted by unforeseen dynamics at the point where an emergent and poorly understood technology meets the implacable friction of the everyday.

Page 309

Whenever we get swept up in the self-reinforcing momentum and seductive logic of some new technology, we forget to ask what else it might be doing, how else it might be working, and who ultimately benefits most from its appearance. Why time has been diced into the segments between notifications, why we feel so inadequate to the parade of images that reach us through our devices, just why it is that we feel so often feel hollow and spent. What might connect our choices and the processes that are stripping the planet, filthing the atmosphere, and impoverishing human and nonhuman lives beyond number. Whether and in what way our actions might be laying the groundwork for an oppression that is grimmer yet and still more total. And finally we forget to ask whether, in our aspiration to overcome the human, we are discarding a gift we already have at hand and barely know what to do with.

Page 313

We don’t even speak of progress any longer, but rather of “innovation.”