Fast & Curious: Short Takes on Random Things

Editor’s Note: I had to make a trip to the Annapolis Valley on short notice yesterday and thought I’d simply be able to post Fast & Curious — which I’d written on Thursday — from the road. But time and circumstance worked against me, which is why you are seeing Fast & Curious on Saturday. I apologize for the inconvenience. I will have to develop some sort of Spectator mobile unit for future such situations.


Tech news

I picked up two interesting and actually, related, pieces of tech information this week and rather than try to work them into a larger piece I thought I’d turn them into two Fast & Curious items, ’cause that’s what Fast & Curious is for. (Apologies to those of you who already know both these things; please feel free to talk amongst yourselves.)

CaptchaThe first concerns those annoying Captcha tests we’re asked to do to prove we’re not robots before accessing our online accounts — you know the drill, “choose all the images of tractors/bicycles/boats/crosswalks/trains/traffic lights/etc.” Their full name is “Completely Automated Public Turing Test to Tell Computers and Humans Apart” and they are not simply intended to establish our bona fides as human beings, they’re also being used to develop the technology for Google’s driverless cars.

I first heard this factoid as a throwaway reference on a podcast and it sent me to the interwebs in search of confirmation. The most detailed account I found came from reporter Shweta Ganjoo on the India Today website, who not only verified the information, but expored it in some detail. It seems that Captcha was introduced in the early 2000s by researchers at Carnegie Mellon University to “filter spam bots that were pretending to be real people.”

The idea was that robots would not be able to read a bunch of letters and numbers that were printed in a distorted format, but humans would. Bots are good at logic. Humans are good at abstract, and printed words and numbers, or in other words images, can be considered abstract. So researchers devised a program that would display a bunch of garbled codes for humans to type in and gain access to a website.

In 2006, the lead researcher on the project, Luis von Ahn, had the idea of using Captcha for “deciphering old smeared text in archival texts” and “reCaptcha” was born. His first project was digitizing the New York Times, as the paper itself explained in 2011:

Dr. von Ahn’s group estimated that humans around the world decode at least 200 million Captchas per day, at 10 seconds per Captcha. This works out to about 500,000 hours per day — a lot of applied brainpower being spent on what Dr. von Ahn regards as a fundamentally mindless exercise.

“So we asked, ‘Can we do something useful with this time?’” Dr. von Ahn recalled in a telephone interview. Instead of making Captchas out of random words printed in a woozy way, why not ask Web users to translate problem words from archival texts?

I’ll let Shweta Ganjoo from India Today take it from here:

So, from 2006 to 2009, which is when Ahn’s start-up was acquired by Google, when millions of internet users across the globe were solving recaptcha codes, they were in a way working for the New York Times and helping it decipher the old smeared text — all for free!

Google next harnessed this free workforce to help with its massive book-digitizing project (a story in itself).

In 2012, Google began to include images in its reCaptcha codes and in 2014, it introduced what Ganjoo calls “a modified version” of reCaptcha called no captcha reCaptcha. (Are you still with me? Because I’m not sure I’m following this myself.) Writes Ganjoo:

Waymo autonomous carThese images were essentially random images and internet users were asked to identify similar looking images, say images of a cat or images of a tree, to prove that they were real people.

In the following years, with Google’s driverless cars becoming an important project for the company, the ‘no captcha ReCaptcha’ gradually started including image-based puzzles that would help AI and machine learning systems behind these cars to learn more about the roads on which would be driving.

Or as Ganjoo puts it, with somewhat startling cheerfulness:

When Google’s driverless cars roll out on busy streets after their ongoing trials end, and when they don’t kill a pedestrian, or hit a cyclist, or run over a fire hose, you can thank yourself. That is because you have taught — and you are teaching — these Google cars now called Waymo cars, to see the roads and everything on them.

I mentioned at the outset that I first heard this link between Captcha and driverless vehicles on a podcast. What I didn’t say was that the podcast host was talking about how automated technology in many areas is nowhere near ready for prime time. He pointed out that we’ve supposedly been helping make Google’s driverless cars smarter for seven years now, and yet, autonomous vehicles seem no closer to becoming a reality now than they were in 2014. In fact, just last week, the CEO of Waymo, John Krafcik, stepped down, prompting CNBC to report:

Krafcik has overseen the company’s biggest milestones, its rebranding to Waymo, partnerships and raised outside funding all while leading enthusiasm through the ranks. But Krafcik’s departure signals a long and arduous reality check to early hype and hope of scaling self-driving vehicles.

“If you look at the past year and a half — there’s been a growing realization within almost all the companies in autonomous vehicle development that this is a much harder problem than we thought,” Sam Abuelsamid, principal analyst at Guidehouse Insights told CNBC Friday. “It wasn’t that long ago people were projecting we’d have robotaxis everywhere by 2020. That hasn’t panned out quite, clearly.”

But while autonomous cars may remain a thing of the future, other types of automation are not only here, they seem poised to increase. (Cue next item.)



That same podcast host also suggested that the purpose of self-checkout kiosks may not actually be to replace all the human cashiers because companies like Walmart are not entirely suicidal — they realize that if all workers are replaced by machines, there will be no market for their goods.

The purpose of the self-checkout kiosk, this theory goes, is to keep workers in line by reminding them how easily they could be automated out of a job. I had never thought of it this way before, so I started looking into just how automation is being used at Walmart and I discovered that, with the rise of online shopping during the pandemic, the company’s current focus seems to be on automating the fulfillment of online orders.

Self Checkout Tesco, Poland

Self Checkout Tesco, Poland (Photo by, Public domain, via Wikimedia Commons)


Writing in mdm — a very niche publication focused on “competitive intelligence for wholesale distribution” — Mike Robuck said that Walmart is “ramping up the build out of mini, automated fulfillment centers” in stores across the United States. A local fulfillment center (LFC):

…can store thousands of items that it knows are in the most demand. Automated bots pull the items off of a shopper’s list from within the fulfillment center before another robot sends them to a workstation where they can be assembled and bagged prior to delivery.

Walmart can augment some products, such as canned goods and electronics, with personnel shoppers that handpick fresh items, such as produce or meat.

“Our automation plan is now ready to scale,” Walmart CEO Doug McMillon said during last month’s earnings call. “We will be investing in our distribution centers, our e-commerce fulfillment centers and in-market fulfillment centers, which will in many cases be inside of or built beside our stores.”

McMillon said they’ve been focusing on pickup for a few years because of “how large the country is” and “how people like to drive their cars,” whereas on the automation side, he thinks they’re “going as quickly and as aggressively as we can and should go. These things will take some time.”

The picture of humans adding fruit and meat to already-packed delivery boxes put me in mind of what the author Emily Guendelsberger said about modern jobs in her excellent book, On the Clock, namely, that while waiting for machines capable of doing everything their employees can do, many employers are using technology to force humans to work like machines. Think trackers monitoring every step and moment of downtime taken by warehouse workers and delivery drivers; quotas that can only be met if workers skip bathroom breaks and pee in bottles; computer-generated work schedules that can force the same worker to close a fast-food restaurant at 2 AM and then turn up at 8 AM for their next shift. (In passing, I think much of the drive to unionize the Amazon warehouse in Bessemer, Alabama is based on the inhumane nature of the work rather than the level of pay — on which note, fingers crossed for the union vote in Bessemer.)

I am still trying to get my head around the issue of automation — is it coming for all our jobs tomorrow or has the threat been wildly overblown? But what I am clear about is that the reality of automation isn’t living up to the promise — which was going to be shorter work weeks, more leisure time and a better quality of life for workers.



Speaking of theories (I was, remember? The “theory” that self checkouts aren’t primarily intended to replace humans?) I have one of my own I want to run past you.

As of Thursday, Nova Scotia had administered 129,809 doses of COVID-19 vaccine and 30,400 people had received two doses.

That means only 3.1% of Nova Scotians are fully vaccinated while 13.3% of the population has had at least one dose. (On the brighter side, even a single dose of vaccine seems to provide significant protection.

According to the CBC’s vaccine tracker from Thursday, while we’re dead last in terms of overall percentage of the population vaccinated, we have a larger percentage of people who are fully vaccinated than does Ontario, B.C., Newfoundland, New Brunswick or Alberta:

Canada vaccine tracker 2021.04.08

I’ve seen a lot of criticism of Nova Scotia’s vaccine rollout, but as Tim Bousquet explained back in January:

In these early months, the province is expected to receive just a tiny number of vaccine doses — if all goes well, 140,000 through April. It’s only after vaccine production ramps up that Nova Scotia is expected to receive very large numbers of doses — over a million over a three-month period beginning in mid- to late-April.

So we should be seeing the rate of vaccination speed up very soon, and that will be a good thing.

But here’s my theory: while rapid and effective vaccine delivery is, of course, something we must strive for (and hold authorities responsible for), the ability to keep a disease at bay via public health measures is a weapon we will be able to use against any future plagues — which may or may not be true of the current COVID-19 vaccines.

I will be as relieved as anyone to be vaccinated, and I’m holding my breath about the chances of one of the variants getting a toehold before the bulk of the population is, but I remain impressed by the way the Atlantic Provinces have handled this disease.

Don’t make me eat my words, Nova Scotia…