Robot lawyers, human cashiers

Paul Raven @ 30-09-2011

Looks like the deeper implications of ubiquitous robotic automation is really starting to sink in. Sure, wondering about the fate of Foxconn’s underpaid manufacturing drones is one thing, but when white-collar professions are threatened, you can believe that handwringing’s gonna happen. That said, Slate charmed me with their subheadline about expert systems and law: “Software could kill lawyers. Why that’s good for everyone else.” What,that needs to be explained?

Oh, I kid, I kid. Not all lawyers are unprincipled scumbags! But as the Slate piece points out, the ones who are could find their business models drying up, especially in the lucrative patent and IP law sectors…

In the last few years, the law has seen a rush of technological innovation, all stemming from computers’ increasing capacity to decipher and understand written documents. Many law firms now use “e-discovery” tools that can scan large caches of evidence in search of interesting facts and figures. Firms also have software to draft legal documents in a fraction of the time a human would take. And a few services on the horizon might do even more—negotiate the terms of a contract, for instance, or determine whether or not you should sue.

Automation will bring legal services to the masses. Many people who ought to hire an attorney to handle business or personal disputes can’t afford to do so. Software could potentially step in when you want to fight your mortgage lender, draw up contracts to start a small business, or sue for child-support payments.

While legal automation will be a boon for those who can’t afford representation, it’s bad news for lawyers. The industry is already in a slump, and law school is no longer seen as a sure path to riches. Because software will allow fewer lawyers to do a lot more work, it’s sure to drive down both price and demand.

A world with less patent trolls and ambulance-chasers sounds just fine to me.

Kevin Kelly’s had his thinking cap on, too; the 7 stages of Robot Replacement are the result:

  • A robot/computer cannot possibly do what I do.
  • OK, it can do a lot, but it can’t do everything I do.
  • OK, it can do everything I do, except it needs me when it breaks down, which is often.
  • OK, it operates without failure, but I need to train it for new tasks.
  • Whew, that was a job that no human was meant to do, but what about me?
  • My new job is more fun and pays more now that robots/computers are doing my old job.
  • I am so glad a robot cannot possibly do what I do.

Zing. First they came for the Foxconn drones, but I said nothing…

… but wait a second, because not all tasks of human productivity are amenable to automation. Furthermore, as end-users, we’re sometimes not very keen on robots that do certain jobs, even if they do they in ways that are technically more efficient. Alex Knapp brings news that will please anyone who hates those self-service tills at the supermarket:

Another supermarket chain is ending self-checkout lanes in favor of more cashiers. This time it’s the Big Ychain, which is eliminating self-checkout in all of its 61 stores because of an internal study that showed that they caused more delays and customers were less satisfied with them than checkout lanes run by lowly humans.

This announcement comes on the heels of major grocery chain Albertson’s announcement in July that it was eliminating self-checkouts in about a third of their stores.

According to a recent Food Marketing Institute study, fewer people are using self-checkouts at the grocery store. They accounted for 22% of all supermarket transactions in 2007, but have since declined to 16% of transactions in 2010. The same study noted that customers were more satisfied with human-run checkout lanes.

Those self-service tills suffer from a number of problems, not least of which are the powerful duo of poor UI design and user stupidity, but deep down, people just don’t like them. Perhaps this is just a lingering sense of the uncanny; they’re still new enough that almost everyone can remember shopping “the old way”, so maybe a sort of lingering cognitive dissonance is at work, which would perhaps fade after a certain acclimatisation period.

But perhaps not. We’re social beings, us humans, and by social I mean the exact opposite of Facebook et al; with a few exceptions, we tend to like interacting with other human beings, and what could be more human, more intrinsic to our cultural bedrock, than the exchange of goods? I’m not suggesting here that a conversation with a checkout operator is likely to be the highlight of your day (in fact, I suspect the subtle reinforcement of social hierarchy that comes from being served may play a role in the deep appeal of such transactions), but I don’t think it’s a wild theory to suggest that the classic “cubicle dystopia” of a world full of people who only ever interact via distance-spanning media is an impossibility. (For one thing, the nigh-universal revulsion we have for the concept – y’know, the thing that makes it a dystopia rather than a utopia –  is a pretty good indicator that, whether biologically or culturally, we’re wired to find that set-up extremely unappealing.)

I’m put in mind of Iain M Banks’ post-scarcity civilisation, The Culture. I can’t remember which novel it appears in, but someone from beyond the Culture is being shown around one of the big arcology/spaceships, and all the restaurants and bars and entertainment venues are staffed by actual living beings (and drones, which as high-functioning AIs, count as people). The visitor expresses surprise that anyone would work when they didn’t have to, and Banks has their Culture host explain that people serve drinks and cook food and play music for others because it’s inherently satisfying to do so. We flinch from the idea at first, but that’s because we’re caught in a world where work gets exchanged for tokens, which are then taken away from us again in exchange for the things we need to survive; in a culture with no money and no physical wants, working for the sheer pleasure of having something to do doesn’t seem crazy at all. Or it certainly doesn’t to me; heck, most of the really horrible jobs I’ve had were horrible because of the conditions and the hours rather than the work itself. That said, I do not include my week working on a waste-collection lorry in that set; there are definitely jobs that are very amenable to automation. (I note wryly that most of them are the ones that are very poorly paid and farmed out to the least fortunate under the current set-up…)

So my theory is that, if all goes well, we’ll automate only the jobs that no one wants to do, but I also suggest that, if the trend is allowed to work out (i.e. no civilisational collapse interferes with our potential trajectory out of the chrysalis of consumerist capitalism), we might find ourselves surprised at which jobs get automated and which ones get kept on. That said, the route between that state and our status quo is a pretty perilous one, and – as usual – it’s the folk at the bottom of the pyramid who’ll be sidelined by automation in a world where we don’t guarantee a universal basic standard of living. It’s high time we faced up to the fact that those two problems are intimately related to one another.


Do nothing, get paid: the “Michelangelos of work avoidance”

Paul Raven @ 09-12-2010

Interesting piece at Forbes here about a sort of person we’ve all known – or maybe even been, to a greater or lesser extent* – in our working careers: they are the “Michelangelos of work avoidance” [via BigThink]:

Work-avoidance Michelangelos know how to stay idle while suffering no consequences or, in some cases, even getting promoted. June lasted in her job for more than a decade before finally being laid off, and when her termination came it had little to do with her lack of productivity. The office was automating her job.

One of her skills was spending little time at her desk or anywhere near the department where she supposedly worked, so that her bosses didn’t even think about her much. Out of sight, out of mind, you might say. “If people don’t think of you, they can’t give you work,” Abrahamson says. Other ways to accomplish that: Arrive at different, unpredictable times of day. Work from home. Set up your schedule so that you frequently change locations.

[…]

If your boss does manage to track you down and try to give you some work, you can strategically deploy a kind of good-natured cluelessness. “The principal here is that you try to give work to a person and come to the conclusion that they can’t even understand the instructions,” Abrahamson explains. In such a case most bosses will figure it’s easier to do the work themselves.

If you perform a specialized function within your office, you can distort the time it takes to get it done. Among June’s supposed jobs was keeping time sheets for her department’s staff. No one else knew the system she’d set up or how long keeping the data took. Thus she could make a task that took minutes appear to consume hours of toil. People with computer expertise who work among Luddites can easily exploit this tactic.

Then there’s what Abrahamson calls the anticipatory screw-up. Make it clear to your boss, in the most pleasant way possible, that you will fail at the assignment she wants to give you. “You don’t have to fail,” advises Abrahamson. “You just have to be clear that you’re going to fail.” Most smart bosses will then give the job to someone else.

Not really surprising that as complex a system as corporate capitalism should have provided niches for freeloaders, nor that human beings – with their innate gravitation towards maximum rewards for minimum outlay – should have taken to them so successfully.

And before you think I’m beating on capitalism alone, I’m pretty convinced that this sort of behaviour goes all the way back into the dawn of history; I suspect shamanism may have arisen due to one or two people per tribe being smart enough to see a way to game the social system. (“Sorry, can’t go hunting with you this week, guys; waaaaaay too busy communing with the gods. But they did tell me that y’all might want to try your luck beyond the third hill to the North…”)

[ * Not me, obviously; I was always a model employee wherever I worked before becoming freelance. SRSLY. ]


The golden age of egalitarianism is twelve

Paul Raven @ 03-06-2010

Ars Technica reports on a research paper that suggests children abandon egalitarianism and develop meritocratic attitudes as they enter adolescence:

… they set up a game where teams of children in grades five through twelve were given access to two websites. One allowed them to earn credits by completing tasks, an analogue of hard work, and the other let them play games or watch videos. At the end of each round, each pair of kids was awarded either 8¢ or 4¢ for each credit he earned, a measure that introduced an element of luck.

The researchers then added the pairs’ earnings together, and asked one of the kids to divide the  money between them. They found that kids in fifth grade tended to be more egalitarian, dividing the money equally regardless of how much either had earned.

However, meritocratic views began to sneak in as the subjects’ age increased: older adolescents not only gave partners less money for earning less, but also awarded themselves less when they had not been as productive. There was also a small but steady number of what the researchers termed  “libertarian” subjects, who divided the money exactly as it had been paid out, without considering the aspect of luck.

Interesting results. However, the real question here is an old favourite: nature or nurture? Is the emergence of meritocratic thinking hardwired into our development, or is it learned socially, from family, friends and media? Plenty of opportunity for further research, methinks… would these results pan out in societies with notably different attitudes to wealth and property ownership?


Multitasking: You can’t do it, my friends

Tom Marcinko @ 03-10-2008

babbage

It might seem like a strange thing to say, coming from a person who’s drinking coffee, answering office email, listening to Juliana Hatfield’s great new album How to Walk Away which I really recommend, and blogging, but multitasking is just about impossible, according to MRI experiments.

…[A] man lying inside the scanner would be performing different tasks, depending on the color of two numbers he sees on a screen. … [W[hen the man in the scanner sees green, his brain has to pause before responding — to round up all the information it has about the green task. When the man sees red, his brain pauses again — to push aside information about the green task and replace it with information about the red task. If the tasks were simpler, they might not require this sort of full-throttle switching. But, [U. Michigan neuroscientist Daniel] Weissman said, even simple tasks can overwhelm the brain when we try to do several at once.

Modern life expects us to do more and more things more quickly, if not simultaneously. If that’s not even possible, at what point do we reorder our tasks and expectations? How will your Bartleby-like character cope?

[Charles Babbage’s brain by Gaetan Lee]


Singularity watch: Vinge on the future

Tom Marcinko @ 26-08-2008

raptureThe New York Times has a brief, appreciative item about Vernor Vinge and his novel Rainbows End. Here’s a nice if-this-goes-on snip:

“These people in ‘Rainbows End’ have the attention span of a butterfly,” [Vinge] said. “They’ll alight on a topic, use it in a particular way and then they’re on to something else. Right now people worry that we don’t have lifetime employment anymore. How extreme could that get? I could imagine a world where everything is piecework and the piece duration is less than a minute.”

[Image: cloudsoup]


Next Page »