2021-01-14

Photo by Sarah Kilian on Unsplash

As 2021 is not too cluster-packed with shitstorms (yet), I’ll go through some of my pending backlog for the last year, together with some fresh catch! I’ve slightly shifted the formatting on these posts as you’ll see below (bye bullet points!).

Wanna feel old and learn some open source history? Remember that this newsletter started as an email, then evolved into codimd, then hopped on to Medium. What happened to codimd though, and why do its links not work atm? Here’s the open source story of Hedgedoc!

Part legend, part just implausible reality. Software folklore contains IT stories to be told around the campfire. No campfire availble? Just run Slack on your laptop. 🔥

This is true shell sorcery in action and I strongly support the poster’s handle.

Revolut seems to maintain its status as one of the most horrifying tech employers in the EU. But fear not — the global stage has strong competitors ;)

(I wanted to put a garbage fire Unsplash image in between these stories, but the only good garbage cans I found where in bright colors, so I had to use them for Google later in the story :/)

What are some of those competitors? Well, let me tell you a story. The story of someone who wrote a warning to Github internal Slack during the storming of the US capitol, then was promptly called over by HR and fired for ‘patterns of behavior’. This started making the rounds both internally, with a memo signed by over 200 coworkers as well as public Twitter, enough that the CEO promised to “look into it”. Shit hit the fan. Turns out this layoff was not a great call, and they, “reversed the separation”. The head of HR took the fall for what was probably not her fault, at least not primarily, and we got some good late-night deleted tweeting.

Speaking of Revolut, over the past months I’ve done a great deal of interviews for my employer, Zivver. One of those was with a former Revolut engineer. When asked about their reasons for switching, the person replied: 
“I believe you may have heard or read about some of the reports of what working in Revolut is like”. 
“Yeah”, I responded, “I’ve seen the Medium articles”. 
“They’re all true”

😰

You may have very adverse reactions to a pixelated uncanny valley Elon Musk joining your Zoom call. I know I would. But a combination of DeepFake and some fast local rendering can work marvels as a party trick I guess (or most probably, end up getting people scammed, blackmailed and killed — sorry for the downer but the DeepNude cases were too real). I’m sorry, but we’re really at the point where people are casually tweeting stuff like this.

This is not so breaking anymore, but amidst the height of the BLM protests in June, Microsoft chose to reject the sale of facial recognition software to police departments. Other vendors have also restricted sales of facial recognition to various groups, whether state-sponsored or not (Hi China!), especially after their wide use in Hong Kong protests. We spoke on the newsletter before about the time when activists walked around capitol with facial recognition cameras on their heads and they were promptly escorted out, which seems almost comical considering recent events. The ironic cherry on top is that they were protesting against the same law that allowed them to legally do that.

But then again, what reason would BLM supporters have to be worried about facial recognition? At least it’s done by computers so it can’t be racist. That’s a thing of the past in 202-oh. Oh shit. Here’s the original tweet btw. We haven’t heard of Google tools acting racist or unethical ever before, right?

Photo by Paweł Czerwiński on Unsplash

While we’re here, feel free to train your personal scanner. Just refresh the page for new results. None of these people exist.

As you probably gather, facial recognition tools and technologies remain quite a troubling and deeply unregulated field. It’s like crypto, except that instead of individual striking back against a system (banking), this one is about a system (state) targetting individuals, which is usually scary. Here’s a John Oliver episode on it, focusing on the shady Clearview.ai (startup that basically provides facial recognition to police agencies based on every public photo ever put on the internet, most of which they have no legal rights on).

On the plus side, DeepFake didn’t destroy society in 2020 as was originally feared, but could it just be that we were too busy? Either way, please don’t destroy AI yet, it’s made my favourite meme.

Title: The Wombats — Let’s Dance to Joy Division!

#articles