Saturday, October 06, 2018

links for the weekend: Amazon pay raise with a trade-off; Russian troll-bots; and report of Chinese seed level hack of hardware

Amazon's going to bump everyone who works for them up to $15 an hour?  There's a trade-off ... via bonuses and stock awards ... Inc. is eliminating monthly bonuses and stock awards for warehouse workers and other hourly employees after the company pledged this week to raise pay to at least $15 an hour.
Warehouse workers for the e-commerce giant in the U.S. were eligible in the past for monthly bonuses that could total hundreds of dollars per month as well as stock awards, said two people familiar with Amazon’s pay policies. The company informed those employees Wednesday that it’s eliminating both of those compensation categories to help pay for the raises, the people said.  ...
To read journalism on bots these days is to get the impression we're at war with Russia, not the kind where Congress officially declares war and we go start bombing places like Odessa or Vladivostock or cities ... but journalists talk about how there's a war against our democracy via the manipulation of systems we've developed for commerce.

I actually really liked Rogue One, myself, and I admit I ended up seeing it just because somebody told me Donnie Yen was in it. The Last Jedi was a trainwreck for scripting reasons, not because there was anything wrong with Kelly Marie Tran's performance as Rose or Rey as the latest unstoppable heroic figure.  Daisy Ridley did just fine with an English language dub of Only Yesterday, one of the more obscure Studio Ghibli releases. Rian Johnson is aware there are fans of the franchise who can't stand The Last Jedi and have articulated why in ways that he can understand.  That's good to know! 

Troll-bot activity is part of the internet.  The axiom that you should basically not read comments at Youtube seems to remain "relatively" relevant.  Besides the fact that so often comments seem virulent and idiotic you may, it turns out, not even be dealing with an actual human so much as with some kind of bot. 

Over at The New Statesmen there's a review of a book that proposes that big tech at the GAFA level (let the reader understand) has undermined traditional democratic processes.

Last month, Apple unveiled the latest version of its watch, featuring new health-monitoring features such as alerts for unusually low or high heart rates, and a way to sense when the wearer has fallen over and, if so, call the emergency services. In itself, that sounds pretty cool, and might even help save lives. But it’s also another nail in the coffin of social solidarity.

Why? Because shortly after the Apple announcement, one of America’s biggest insurance companies, John Hancock, announced it would stop selling traditional life insurance, and would now offer only “interactive” policies that required customers to wear a health-monitoring device – such as an Apple Watch or Fitbit. But such personalised insurance plans undermine the social spreading of risk that makes insurance a public good. Knowing every little dirty secret about our lifestyles, such an insurer will be heavily incentivised to make the riskier customers pay more in premiums than the healthy-livers. Eventually, the fortunate will subsidise the less fortunate to a far smaller degree than they do on traditional insurance models. For those who get sick, this will literally add insult to injury. 
If, practically speaking, you can’t opt out of a health care platform, or switch from the education platform your local school uses, then unaccountable corporate monopolies have usurped the functions of government. Moore calls this “platform democracy”. You might equally suggest it as a new meaning for “technocracy”, which up till now has meant rule by experts. Soon, technocracy might mean rule by people who don’t understand anything, but think that data alone constitutes expertise; people who glory in the “engineering ethos” of rapid prototyping and deployment; or, as Facebook’s old motto had it, “move fast and break things”. This is fine when you are building a trivial app; it’s not so fine if the things you are breaking are people and social institutions.

He begins by bringing the reader up to speed, in lucid detail, on Steve Bannon and the Breitbart website, as well as the story of Cambridge Analytica. He explains what we know about Russian interference in the 2016 US presidential election, while making the important point that such operations are not at all new. During the Cold War, the USSR and its puppet regimes ran energetic fake-news operations against the West. The only difference now is that modern technology makes disinformation operations much more effective, as falsehoods can go viral around the globe in a matter of minutes. Putin now has his own social-media sock-puppet farm, hidden in plain sight under the bland name of the “Internet Research Agency”. (It does about as much research as Jacob Rees-Mogg’s “European Research Group” for hard Brexiteers.)

This leads directly into Moore’s larger argument, which is that for reasons of profit the tech platforms actively turned themselves into machines perfectly suited to the dissemination of anarcho-nationalist hatred and untruth. Until recently, Moore notes, Facebook rarely thought about politics, and if it did “it tended to assume the platform was by its nature democratising”. But ahead of its 2012 stock-market floating, it went “all out to create an intelligent, scalable, global, targeted advertising machine” that gave advertisers granular access to users. And so it created the most efficient delivery system for targeted political propaganda the world had ever seen.

It wasn’t just the bad guys who noticed this. In 2012, Barack Obama’s blog director Sam Graham-Felsen enthused: “If you can figure out how to leverage the power of friendship, that opens up incredible possibilities.” The possibilities that Facebook has since opened up would have seemed incredible six years ago. A member of the Trump campaign team openly described one aspect of their Facebook campaign as “voter suppression operations” aimed at Democrats, using something called “dark posts”. These allowed operators to conduct sophisticated testing comparing the effects of different kinds of adverts, creating, as Moore puts it, “a remarkably sophisticated behavioural response propaganda system”

Technocracy doesn't always have to be by way of direct rule.  Let's put this another way, if in the past technocracy might have meant that the technocrats had meetings in which things got decided perhaps there's a new layer of distancing that's possible, by way of app development or protocol implementation.  Formulas, protocols and data-mining can be used. 

The idea that Obama wasn't part of "the bad guys" ... not so sure about that.  If the difference now is not that the USSR no longer exists but that Russia can spread fake news faster by dint of the exponentially greater power to have things go "viral" then the difference wouldn't be the evil Russians but that "we" created a network of communications systems and marketing platforms that exponentially boosted the power "they" have in making the same old fake news.  Is that something to put on the Russians or a potential indictment of our own power to multiply by orders of magnitude our own stupid?  It's starting to seem more and more like the latter, not that you should care what a blogger writing on the weekend thinks ... .

The review closes with an observation that Estonia has an information silo approach to citizen data that currently precludes the kind of consolidations that have become normative in the West.  I've got too much reading on my plate as it is but that's an interesting case study for a counter to what the norm seems to have been in the United States.  Thus another part of links for the weekend.

In the midst of writing about Russia it could seem as though the subject of China gets skipped altogether.  Whether or not Russia successfully hacked the 2016 election there are other things to be concerned about.  Was the giant break of data from the OPM Russian caused?  Don't honestly recall that it was.  More in the news in the last week is Chinese hacking.

The ramifications of the attack continue to play out. The Trump administration has made computer and networking hardware, including motherboards, a focus of its latest round of trade sanctions against China, and White House officials have made it clear they think companies will begin shifting their supply chains to other countries as a result. Such a shift might assuage officials who have been warning for years about the security of the supply chain—even though they’ve never disclosed a major reason for their concerns.
One Friday in late September 2015, President Barack Obama and Chinese President Xi Jinping appeared together at the White House for an hourlong press conference headlined by a landmark deal on cybersecurity. After months of negotiations, the U.S. had extracted from China a grand promise: It would no longer support the theft by hackers of U.S. intellectual property to benefit Chinese companies. Left out of those pronouncements, according to a person familiar with discussions among senior officials across the U.S. government, was the White House’s deep concern that China was willing to offer this concession because it was already developing far more advanced and surreptitious forms of hacking founded on its near monopoly of the technology supply chain.

In the weeks after the agreement was announced, the U.S. government quietly raised the alarm with several dozen tech executives and investors at a small, invite-only meeting in McLean, Va., organized by the Pentagon. According to someone who was present, Defense Department officials briefed the technologists on a recent attack and asked them to think about creating commercial products that could detect hardware implants. Attendees weren’t told the name of the hardware maker involved, but it was clear to at least some in the room that it was Supermicro, the person says.

The problem under discussion wasn’t just technological. It spoke to decisions made decades ago to send advanced production work to Southeast Asia. In the intervening years, low-cost Chinese manufacturing had come to underpin the business models of many of America’s largest technology companies. Early on, Apple, for instance, made many of its most sophisticated electronics domestically. Then in 1992, it closed a state-of-the-art plant for motherboard and computer assembly in Fremont, Calif., and sent much of that work overseas.

Over the decades, the security of the supply chain became an article of faith despite repeated warnings by Western officials. A belief formed that China was unlikely to jeopardize its position as workshop to the world by letting its spies meddle in its factories. That left the decision about where to build commercial systems resting largely on where capacity was greatest and cheapest. “You end up with a classic Satan’s bargain,” one former U.S. official says. “You can have less supply than you want and guarantee it’s secure, or you can have the supply you need, but there will be risk. Every organization has accepted the second proposition.”

So in the decades in which the United States outsourced manufacturing to Asia we ended up discovering that by doing so companies that sought to increase profits and cut down manufacturing costs managed to do that at the price of foreign governments and military branches figuring out that they could build the power to hack and spy on us into the manufacturing processes we outsourced?  It vaguely reminds me that I was blogging now and then about how the odds of formal war were remote when so many other options of doing battle at the level of our commerce and data infrastructure were available.   If the axiom in military history is that too many nation-states make the mistake of fighting the last war they fought then it might seem as though the problem we're running into is that wars are no longer being fought at the level of soldiers in trenches when they can be fought at the level of weaponizing our consumption habits against us. 

On a relatively lighter note, apparently small university level presses are having a moment.


chris e said...

There are a number of things off about that bloomberg story - a number of details that don't make sense taken together, along with the strong denials from the parties concerned (as opposed to non-denial denials - which you'd expect if they were trying to hedge).

On the other hand, if it means that people look more carefully at their supply chain, then that's probably not a bad end result.

Wenatchee the Hatchet said...

whether it's the China threat or the Russia threat it's hard to shake the sense that "yes" some kind of shady somethings have been going on but having not been very zealous for Gulf War 2 in the last nearly 20 years I just don't get the sense that now's the time to ramp up tensions with either China or Russia.

I mean, I guess if someone wanted to make a case for a restricted trade set up or trade war this is the kind of coverage that could ensure that would be done and that might be all that matters for trade hawks or other sorts of hawks.

My own hobby horse, having worked dead end jobs my adult life, is that if we shipped off a lot of the manufacturing side of things overseas in the last thirty years this kind of thing would seem to be the trade off. I doubt that trying to tariff/protectionist the US back into making these things in the U.S. would work well, though.

The more I thought about it the more the ultra tiny chip being able to do so much seemed a bit sketchy.

Being pickier about the supply chain might be the Captain Obvious point to make but it's always a good point. :)