Inside Amazon

An interesting counterpoint to the undercover “stings” which support the narrative that Amazon exploits its workers, written by an actual Amazon warehouse worker.

While warehouse work doesn’t sound particularly appealing or financially rewarding to your average white-collar worker, he makes the point that Amazon’s are pretty decent warehouses to work at and the pay is quite a bit better than other jobs on offer. Sure, $15/hr isn’t much, but it’s 36% better than $11/hr elsewhere.

Bloodworth says he worked 10-and-a-half-hour days at Amazon, which sounds pretty brutal. Maybe they do things differently in the UK, but my Amazon sortation center is very flexible about the hours it offers. When I was applying for the job online, Amazon allowed me to create a schedule tailored to my needs. They asked me how many hours a week I’d like to work, and which days of the week suited me best. They asked if I preferred to work evenings, overnight, early mornings, days. After compiling this info, they gave me a shift that fits me like a glove. I work four and a half hours a day, five days a week. My shift begins at 6:30 am and ends at 11 am. But, when I need a bit more money, I can go to work at 5 am and pick up an extra 90 minutes of work pretty much whenever I want. Many of my co-workers add hours to their days whenever they are in need of a little extra cash, and we can take voluntary unpaid time off just about whenever we like.

Privacy in the Information Age

Idle Words has a good article on [The New Wilderness](https://idlewords.com/2019/06/the_new_wilderness.htm} on the nature of privacy in the Information Age. What does it mean when your every move is tracked and recorded online, and increasingly offline as well?

Until recently, ambient privacy was a simple fact of life. Recording something for posterity required making special arrangements, and most of our shared experience of the past was filtered through the attenuating haze of human memory. Even police states like East Germany, where one in seven citizens was an informer, were not able to keep tabs on their entire population. Today computers have given us that power. Authoritarian states like China and Saudi Arabia are using this newfound capacity as a tool of social control. Here in the United States, we’re using it to show ads. But the infrastructure of total surveillance is everywhere the same, and everywhere being deployed at scale.

The author discusses the similarity with the growth of environmental regulation as mankind changed from a being part of Nature to being a threat instead and suggests that we need to start thinking of global regulations around what is and is not acceptable when it comes to mass surveillance.

We’re at the point where we need a similar shift in perspective in our privacy law. The infrastructure of mass surveillance is too complex, and the tech oligopoly too powerful, to make it meaningful to talk about individual consent. Even experts don’t have a full picture of the surveillance economy, in part because its beneficiaries are so secretive, and in part because the whole system is in flux. Telling people that they own their data, and should decide what to do with it, is just another way of disempowering them.

Our discourse around privacy needs to expand to address foundational questions about the role of automation: To what extent is living in a surveillance-saturated world compatible with pluralism and democracy? What are the consequences of raising a generation of children whose every action feeds into a corporate database? What does it mean to be manipulated from an early age by machine learning algorithms that adaptively learn to shape our behavior?

Uber: Unicorn or Ponzi Scheme

Interesting read from American Affairs, delving into Uber’s pre-IPO financials and making the case that it’s not the great revolution in transport it claims to be admin a way resembles a Ponzi scheme with earlier investors making massive profits from the suckers lured in later.

Uber’s investors, however, never expected that their returns would come from superior efficiency in competitive markets. Uber pursued a “growth at all costs” strategy financed by a staggering $20 billion in investor funding. This funding subsidized fares and service levels that could not be matched by incumbents who had to cover costs out of actual passenger fares. Uber’s massive subsidies were explicitly anticompetitive—and are ultimately unsustainable—but they made the company enormously popular with passengers who enjoyed not having to pay the full cost of their service.

Uber’s financials don’t tell a great story…

Uber’s GAAP profit margin was –135 percent in 2015. It appeared to improve to –51 percent in 2017 and (adjusting for the divestiture and noncash equity gains discussed above) –35 percent in 2018. Yet these subsequent “improvements” were not driven by efficiency gains, but by the ability to force driver take-home pay down to minimum wage levels. If Uber drivers still received their 2015 share of each passenger dollar, Uber’s negative margins would still be in the triple digits.

I wonder how many of today’s unicorns are where they are today due to the essentially zero cost of money since the GFC supporting otherwise insupportable business models, or at least permitting those unprofitable business models to persist for far longer than would otherwise be the case. Tesla’s another candidate. Great cars by all accounts, and Musk is a great entrepreneur, but making the transition to established, profitable car maker seems perpetually out of reach.

Apple Car Common Sense

A lot of common sense from Monday Notes on what the rumours related to an Apple Car could actually refer to and what Apple could, or could not, be up to.

Reality quickly kills the warm feeling. There’s trouble with the autonomous part of the dream: Sober people (see last two weeks’ Monday Notes) agree that full “Level 5” automation — no need ever for human intervention, from arbitrary point A to point B, in any weather — is decades away. Like anyone else’s EV the Apple Car will feature a mix of driver assistance services with no clear way to get to full autonomy, particularly when compared to Waymo and Tesla with their millions of test miles and mountains of real world data.