More on “grim” automation – Notebook

A few weeks back my book review of two “the robots are taking over” came out over on The New Stack. Here’s some responses, and also some highlights from a McKinsey piece on automation.

Don’t call it “automation”

From John Allspaw:

There is much more to this topic. Nick Carr’s book, The Glass Cage, has a different perspective. The ramifications of new technology (don’t call it automation) are notoriously difficult to predict, and what we think are forgone conclusions (unemployment of truck drivers even though the tech for self-driving cars needs to see much more diversity of conditions before it can get to the 99%+ accuracy) are not.

Lisanne Bainbridge in her seminal 1983 paper outlines what is still true today.

From that paper:

This paper suggests that the increased interest in human factors among engineers reflects the irony that the more advanced a control system is, so the more crucial may be the contribution of the human operator.

When things go wrong, humans are needed:

To take over and stabilize the process requires manual control skills, to diagnose the fault as a basis for shut down or recovery requires cognitive skills.

But their skills may have deteriorated:

Unfortunately, physical skills deteriorate when they are not used, particularly the refinements of gain and timing. This means that a formerly experienced operator who has been monitoring an automated process may now be an inexperienced one. If he takes over he may set the process into oscillation. He may have to wait for feedback, rather than controlling by open-loop, and it will be difficult for him to interpret whether the feedback shows that there is something wrong with the system or more simply that he has misjudged his control action.

There’s a good case made for not only the need for humans, but to keep humans fully trained and involved in the process to handle errors states.

Hiring not abating

Vinnie, the author of one of the books I reviewed, left a comment on the review, noting:

For the book, I interviewed practitioners in 50 different work settings – accounting, advertising, manufacturing, garbage collection, wineries etc. Each one of them told me where automation is maturing, where it is not, how expensive it is etc. The litmus test to me is are they stopping the hiring of human talent – and I heard NO over and over again even for jobs for which automation tech has been available for decades – UPC scanners in groceries, ATMs in banking, kiosks and bunch of other tech in postal service. So, instead of panicking about catastrophic job losses we should be taking a more gradualist approach and moving people who do repeated tasks all day long and move them into more creative, dexterous work or moving them to other jobs.

I think Avent’s worry is that the approach won’t be gradual and that, as a society, we won’t be able to change norms, laws, and “work” over fast enough.

McKinsey

As more context, check out this overview of their own study and analysis from a 2015 McKinsey Quarterly article:

The jobs don’t disappear, they change:

Our results to date suggest, first and foremost, that a focus on occupations is misleading. Very few occupations will be automated in their entirety in the near or medium term. Rather, certain activities are more likely to be automated, requiring entire business processes to be transformed, and jobs performed by people to be redefined, much like the bank teller’s job was redefined with the advent of ATMs.

Further:

our research suggests that as many as 45 percent of the activities individuals are paid to perform can be automated by adapting currently demonstrated technologies… fewer than 5 percent of occupations can be entirely automated using current technology. However, about 60 percent of occupations could have 30 percent or more of their constituent activities automated.

Most work is boring:

Capabilities such as creativity and sensing emotions are core to the human experience and also difficult to automate. The amount of time that workers spend on activities requiring these capabilities, though, appears to be surprisingly low. Just 4 percent of the work activities across the US economy require creativity at a median human level of performance. Similarly, only 29 percent of work activities require a median human level of performance in sensing emotion.

So, as Vinnie also suggests, you can automate all that stuff and have people focus on the “creative” things, e.g.:

Financial advisors, for example, might spend less time analyzing clients’ financial situations, and more time understanding their needs and explaining creative options. Interior designers could spend less time taking measurements, developing illustrations, and ordering materials, and more time developing innovative design concepts based on clients’ desires.

“we overestimate our own capacity for truly independent thought”

From The Attention Merchants:

Any communication, Lippmann came to see, is potentially propagandistic, in the sense of propagating a view. For it presents one set of facts, or one perspective, fostering or weakening some “stereotype” held by the mind. It is fair to say, then, that any and all information that one consumes—pays attention to—will have some influence, even if just forcing a reaction. That idea, in turn, has a very radical implication, for it suggests that sometimes we overestimate our own capacity for truly independent thought. In most areas of life, we necessarily rely on others for the presentation of facts and ultimately choose between manufactured alternatives, whether it is our evaluation of a product or a political proposition. And if that is true, in the battle for our attention, there is a particular importance in who gets there first or most often. The only communications truly without influence are those that one learns to ignore or never hears at all; this is why Jacques Ellul argued that it is only the disconnected—rural dwellers or the urban poor—who are truly immune to propaganda, while intellectuals, who read everything, insist on having opinions, and think themselves immune to propaganda are, in fact, easy to manipulate.

I’m hoping to bundle this book up into my next book review for The New Stack.

A false choice: systems of record vs. systems of engagement – Pivotal Conversations #49

What’s the best way to categorize and prioritize your IT projects? Splitting them up between systems of record (ERP) and systems of engagement (user-facing apps) is a popular mode of thinking, highly related to bi-modal IT. In this episode, guest Ian Andrews explains why this framing is a bad idea and offers a value-driven way of thinking about it instead, along with plenty of commentary from Coté and Richard.

Subscribe to the podcast feed, check it out in Soundcloud, or listen below:

Subscribe, follow, feedback

News

Ian Andrews on bimodal, systems of engagement vs. systems of record

  • Systems of engagement vs systems of record – why is this distinction not helpful based on Ian’s conversations with actual customers. And, of course, bimodal.
  • Spring Boot’s story, and Spring Cloud.
  • Contrasting those with JEE needs shifting. That InfoQ piece Ian references.

Life after artisanal pork rinds (i.e. tech M&A), CostCo Down Under – Software Defined Talk #86

This week’s episode goes over several recent M&A deals in tech: AppDynamics, Trello, and Apiary. Plus, Matt Ray’s report from CostCo Australia. Check out the full show notes, subscribe, and/or listen below:

Tech must rethink working with the hobgoblins, cf. scorpions & turtles – TrumpTech

And just today, [Bloomberg is reporting] another executive order being drafted focused on work visas that tech companies depend on, which will have a big impact on how critical talent is recruited. According to the report, “companies would have to try to hire American first and if they recruit foreign workers, priority would be given to the most highly paid.”

More: I was at a chock-full event in Palo Alto last week, as tech types planned their attack on the defunding of Planned Parenthood and the reinstatement of the global gag rule by Trump and the GOP that restricts foreign aid to those organizations that reference abortions in family planning. It was a move that Facebook COO Sheryl Sandberg spoke out against last week. “We don’t have to guess,” she wrote, noting that the move is a disaster for women globally. “We know what this will do.”

What else? Well, now there are rumors that Trump could sign another executive order aimed at restricting advances in rights made by gays and lesbians, such as allowing people to refuse to do business with them due to religious objections (expect federal legislation here too). And, earlier this week, press secretary Sean Spicer also said, “I don’t know,” when asked if Trump would rescind a Barack Obama executive order banning anti-LGBT discrimination by federal contractors.

Given tech leaders have been very vocal in their support of gay issues, which are important to their employees, if Trump does any of this, it should go off like a Roman candle in Silicon Valley.

Kara Swisher lays out the naïve basis of thinking things would go well, then a suggestion for tech black-balling Trump and his cronies: they’ll likely take actions to harm tech labor and culture, and business as well.

I’m all for fail-fast when it’s in the world of business, but the confusion and whip-lashing in the immigration stuff this weekend shows that the hobgoblins are both in over their heads and also don’t realize how dangerous caviler their policy setting is.

My concern – as it always has been – is that the Trump adminstration is both against my morals and inept enough to dangerously goof things up. Their bald-facing (any petty!) lying and moronic roll-our of this Muslim Ban shows them to be unfit.

Just imagine how damaging that ineptitude will be when it comes to the nuanced, complicated world of tech. For example: anyone fancy their electronic toys being 30% to 45% more expensive?

Read more reporting on how poorly this was all handled from today’s NY Times piece.

See also Victor Rozek’s history-injected opinion and advice in this topic. Also, a round-up of tech executives public statements on the topic, from Sam Biddle. Also, RedMonk’s call to action.

Costs that go into a $185 shirt, including snacks & other human affordances

At a fast-fashion retailer such as H&M, a simple cut-and-sew top can cost as little as $15. At Gap, something similar might run about $45. At Elizabeth Suzann, a small fashion label based in Nashville, Tennessee, one of the brand’s minimal kimono-sleeved t-shirts, made of cotton twill, is $140.

This seems like the kind of analysis that’ll be handy in the up-coming trade wars.

Link

Book Review: Automation & tech ethics, book review

These two books go well together because the first describes how automation is lowering the need for labor, leading, likely, to less jobs, while the second provides a compendium of examples of such software-driven labor change.
Vinnie’s book has the optimism of a technologist, while Avent’s is much more fraught. Both accurately describe how IT is optimizing and replacing “analog” labor and businesses, leaving the core problem of devaluing human labor, perhaps to the point of eliminating millions of jobs, permanently. Vinnie’s optimism is the usual believe that we can figure it out, mostly by being more humane in our politics and safety nets, but also in the belief that new problems and jobs will come about. Avent, on the other hand, offers little in the way of solace.
As the review in his magazine, The Economist, put it: “I found the virtuosity with which Mr Avent knocked down possible solutions disquieting.” Aside from actually reading the book, the lecture Avent gave at LSE is good stuff too.
Check out the full review.

Reactive applications, with Josh Long – Coté Show #18

While in San Francisco for a Pivotal meeting, I recorded an episode with Josh Long. We talk about what reactive programming is and why you’d use it. Also, dish towels. Also, check out the livestream of this if you’re into video.

If you haven’t subscribed already, subscribe to the Coté Show. I do little monologs and interviews like this in it, somewhat irregularly.

Avoid the Ninja Anti-Pattern, Planning Out Your Cloud Platform Project – Pivotal Conversations #48

How do containers fit into your cloud native planning? That’s a the question we start with this week, with (returning guest) John Feminella. We quickly arrive at a conversation on the larger question which is how to build a cloud platform and the allure of building it yourself. Also, we cover recent news in the infrastructure software space.

Subscribe, follow, feedback

News