Tim Boucher

Questionable content, possibly linked

Category: Review

Duolingo percent fluent score is way off

I’ve been using Duolingo for several weeks now to create a base floor of knowledge in German. I’ve decided their percent fluency score is pretty much random…

The length of time I’ve been 26% Fluent in German is ridiculous. Obviously, I understand this is a hard thing to measure, and that people need to incentivize their progress as they work through the steps towards a goal like this, but this is one element I really wish they would revisit.

I also don’t really care at all about their virtual reward currency, “lingots,” which allow me to “buy cool stuff at their virtual store.” It has almost no bearing on my usage of the application.

Lastly, don’t fool yourself into thinking that because you’re receiving good marks on a language via Duolingo that this will automatically translate into a “high score” when you go try to practice speaking that language in the actual country or community. It’s a fun app, and can be a good supplement to other learning, but I wouldn’t trust a 100% fluency score off Duolingo to equal real world fluence by a long shot…

Trust-breakers (Chinese Social Credit System)

Legal effects of automated processing, a comparison.

I’ve been reading about China’s emerging social credit system, Sesame Credit.

“The score is used to rank citizens of China based on a variety of factors like loyalty to the Chinese government and loyalty to Chinese brands based on social media interactions and online purchases. The rewards of having a high score include easier access to loans, easier access to jobs and priority during bureaucratic paperwork.”

Here are a couple articles to get you started:

Blah blah blah, obligatory Black Mirror reference. Now that we have that out of the way, from the CNBC link:

“When rules are broken and not rectified in time, you are entered in a list of ‘people subject to enforcement for trust breaking’ and you are denied access from things. Rules broken can lead to companies being unable to issue corporate bonds or individuals not being allowed to become company directors,” Creemers said.

Basically, a bunch of apps and agencies work together to rank your behavior and profile you socially as either a trust-keeper or trust-breaker as described above. Via the FP link:

“By 2020, the government says that social credit will “allow the trustworthy to roam everywhere under heaven while making it hard for the discredited to take a single step.””

I had a dream years ago that I used a urinal at a shopping mall, and the system automatically administered a drug test on me, which I failed. Not being a cell phone user, I needed to then borrow a friend’s phone to make a call and the system linked my voice-print to my biometric/pee test and I was disallowed from using my friend’s phone. Such a unified system may be a few years off still, but the possibility is becoming tantalizingly real. I might even say it’s, on some level of implementation, pretty much inevitable.

I’ve been following a parallel strand of research these past few months. It’s partly intuition, partly investigative leg work, but it’s lead from public records databases used by private eyes 🕵 to the vast store-houses of data kept commercially against named individuals by data brokers. I’m still largely in the dark about how data brokers operate, and, er, broker — despite hours spent around the subject on Youtube. But I have to assume that those storehouses of information about people have to be searchable — at a price. Whether it’s pseudonymized in aggregate, or traceable to an identified or identifiable individual, all this information exists somewhere out there, waiting to be linked up and put to use.

Meanwhile, half a world away, Europe is set to roll out it’s GDPR next May (2018) which will quite possibly make very difficult – or at least very different – such a social credit system were it to be rolled out to customers in the European Economic Area.

I explored this question elsewhere, of processing of personally-identifying information linked to automated decision-making, and profiling, with “legal effects”. So I won’t completely rehash it here, but to quote Article 22 of the GDPR:

“The data subject shall have the right not to be subject to a decision based solely on automated processing, including profiling, which produces legal effects concerning him or her or similarly significantly affects him or her.”

The recital for that article mentions explicitly as an example the “automatic refusal of an online credit application” as something that has a legal effect.

I guess this is worth quoting more extensively from second half of Recital 71:

“Such processing includes ‘profiling’ that consists of any form of automated processing of personal data evaluating the personal aspects relating to a natural person, in particular to analyse or predict aspects concerning the data subject’s performance at work, economic situation, health, personal preferences or interests, reliability or behaviour, location or movements, where it produces legal effects concerning him or her or similarly significantly affects him or her.”

So this is pretty much explicitly describing an all-encompassing “social credit” system such as is currently being live-beta tested on Chinese society. In other words, Europe is baking into their privacy & data protection regime this idea that the fundamental rights of humanity (from which privacy/data prot. are derived) are incompatible with automated decision-making based on data processing with (potentially negative) legal consequences.

That’s huge.

To me, as we move into the Algorithmic Society (and it’s many diverse, fascinating and horrifying forms, instantiations and iterations), this will be a fundamental tension as humanity transitions to greater and greater levels of algorithmic control, automation and governance of day-to-day life.

Quoting from Art. 22, 3:

“…at least the right to obtain human intervention on the part of the controller, to express his or her point of view and to contest the decision.”

The subject still blows my mind. Partly because we now live in (or very soon will) a society where such rules have become necessary. The algorithmic age where Trust Breakers ™ can’t buy train tickets, or make a phone call. But in exchange for keeping your score up, you’re eligible for ultra-fast lane physical access from Boston to DC in 15 minutes, with no control checkpoints, minimal surge pricing, only light deep packet inspection and limited throttling. [See full Terms & Conditions.]

Or a world where slow, boring, crappy, unreliable human bureaucratic decision-making is baked into break-points in societal algorithms to ensure some sort of fairness, humanity, tolerance, resilience (and maybe forgiveness?) into what will otherwise most assuredly become a mesh of AI’s vying for planetary control…

I’m sorry, that’s just where my mind goes when I pull out my 🔮. I think it’s why I like the GDPR as a document in the first place: it reads like a dystopian cyberpunk text that young punks in the future will repeat back verbatim to quasi-governmental robots that are beating the snot out of them because their social credit score has fallen too low.

Whoops, I went off into la-la land there again at the end*. But what can I say? I’m on vacation! 🌴🍹

* You try talking about this without landing on the subject of killer robots. It’s not so easy. It’s like Godwin’s law, but for killer robots and data processing.

Javascript rapid word input tool

I went off on pretty much a tear earlier investigating the possibility of coming up with some kind of rapid communication board which would allow you to input words, not letters.

I went once or twice around the bend, and found the closest match in an app called DocsPlus which gives you the ability to create customizable word-bars. There’s a 28 day free trial. It’s interesting, but my use case is to be able to rapidly paste in the results of these sentence creation actions into Firefox in a spreadsheet. It was too combersome with switching back and forth between tabs to access other word bars.

So I cooked up some Javascript I’m still tinkering with which looks at the moment like this:

Coursera needs quicker access to videos

Dear Coursera,

I’m enjoying the video lessons from Coursera’s “Internet Giants” course, but I find the videos are too hard to get to.

This is the public home page of the course, which is what I always end up accessing through via my web browser autocomplete in the address bar.


Which becomes when you click “Go to Course” (I’m not already at the course?):


And this confusing chart of weeks…

I appreciate that they want to replicate some sort of “classroom experience” which would take place across successive weeks in time, but that’s not how I consume the product. This is the internet after all. I watch two or three of the videos in a row, as time permits. And then I come back at staggered intervals. Every couple of days or so.

So the meta-conception of weeks becomes actually a hindrance for me to simply do the thing that is integral to my use of the product: just watch the next video.

  • I don’t care about the quizzes.
  • I’m not of the belief someone will care if I paid $66 in the future to officially “pass” this online course.
  • I don’t have time for the extra readings, though I’m sure they are fascinating.

I just want to see the next video, and to “ride the wave” of the subject matter as it unfolds through the narrative of a skilled instructor.

How about from the initial homepage, there’s just a button that says Continue?

Soomz webcam cover review

I recently purchased from Amazon.ca, a pack of 3 web cam covers made by Soomz, a Swiss company.

The price I thought, $15 for 3, was a bit high but the product fits on both a Macbook Pro and Air, and seems to “work fine.”

I’m pretty sure, and read as much, that these won’t work that well on a mobile phone or other device which has more in-and-out-of-pocket type action. They are attached by adhesive, so that would probably wear out. But for laptop use they seem great.

I had been covering my webcam for years with black electrical tape, but that is less than ideal if you’re regularly engaging in video conferencing, and having to remove and reapply the tape with any regularity. This is nice because it operates by a simple slider mechanism, and I like the peace of mind it offers by being able to instantly visually verify that your camera is closed.

Internet Giants on Coursera

I’ll admit that I’ve always been skeptical of Coursera, because it just sounds… I don’t know. Cheap or something. But at the same time, I’ve gotten a ton of value out of Khan Academy, so I finally figured Why not? and signed up for Randy Picker’s University of Chicago-linked course, Internet Giants: The Law and Economics of Media Platforms.

Given the ubiquity of social media platforms in our daily lives, this seems like a big deal.

There’s some kind of mechanism with Coursera where I guess certain courses you can pay for have some kind of “proof” you took it, or like this one you can just do it for free. Which is what I’m doing. At first, I was participating in the quizzes and trying to look at the extra readings. But I also happen to have a “job” and some kind of “life,” so I’m just watching the videos in sequence, and they are great.

In actual fact, I wish there were a simple mechanism just to run through the playlist of the videos and pick up where I left off. Instead, you have to hop around through their arbitrary system of “weeks” (which has nothing to do with anything in the fragmented time-space of internet assimilation. But whatever, it’s a minor UX quibble for a product I’m able to access for free with what seems like an increasing amount of quality content. I’ve “enrolled in” but not begun one or two others, and there are quite a number of others which I wish I could bookmark, or put on a list or something and come back to later.

Anyway, highly recommend Randy Picker’s online course for excellent context on what’s going on today.

Page 2 of 2

Powered by WordPress & Theme by Anders Norén