Hey,
Some of you might have spotted that I accidentally wrote 2020 under the title of last weeks email, which whilst wishing you all a happy new year was my one job...
Anyway, C’mere to me…
🇺🇸
Privilege.
So everyone saw what happened in the States on Wednesday night…
One thing that was very evident as the world watched everything unfold was the slow and double-standard reaction from authorities.
Here are two analysis pieces shared online the day after that I thought were interesting if you want to read more on the issue, including one on the first data-backed study comparing protests, of all natures over, the last year and the subsequent actions the police took to deal with them.
“Between May 1 and November 28, 2020, authorities were more than twice as likely to attempt to break up and disperse a left-wing protest
(For the purposes of ACLED’s analysis, this includes protests that are anti-Trump or pro-Biden, in support of the Democratic Party, affiliated with the BLM movement, associated with Antifa or left-leaning groups such as Abolish ICE, the NAACP, the Democratic Socialists of America as well as Count Every Vote demonstrations.)
than a right-wing one.
(Includes pro-Trump or anti-Biden rallies, events in support of the Republican Party, pro-police demonstrations such as Back the Blue or Blue Lives Matter movements that often organize against the BLM movement, demonstrations involving QAnon conspiracy theories, militias or street movements to “keep the peace” during an event, and Stop the Steal demonstrations.)
And in those situations when law enforcement chose to intervene, they were more likely to use force — 34 percent of the time with right-wing protests compared with 51 percent of the time for the left. Given when this data was collected, it predominantly reflects a difference in how police respond to Black Lives Matter, compared with how they respond to anti-mask demonstrations, pro-Trump extremists, QAnon rallies, and militia groups.”
(read more - FiveThirtyEight)
“During the Black Lives Matter protests this past spring, National Guard troops in combat gear stood on the steps of the Lincoln Memorial three lines deep. Around the same time, U.S. Park Police tear-gassed nonviolent protesters in Lafayette Square, in Washington, D.C. The Capitol Police made more arrests on each of the first three days of the confirmation hearings for Supreme Court Justice Brett Kavanaugh, in September, 2018, than they did Wednesday.
-
The invaders may be full of contempt for a system that they think doesn’t represent them, but on Wednesday they managed to prove that it does. The system, which shrugged off their violence like it had been a toddler’s tantrum, represents them. It’s the rest of us it’s failing to protect.”
(read more - The New Yorker)
One hopeful thing, albeit there is still so much to left to question, is that law and order did somewhat prevail.
Image - Reddit
Tech
I am not a Robot ☑
The CAPTHCA
A few weeks ago I shared some acronyms, one being CAPTCHA - Completely Automated Public Turning test to tell Computers and Humans Apart.
One form of a CAPTCHA, usually found on a form you submit to a website, is that box you sometimes have to tick to confirm that you are in fact not a robot…
Spoiler alert, it’s purpose has nothing to do with ticking the box - as many memes and TikToks have demonstrated it is possible to get a machine to tick the box successfully.
I’ll explain the concept of a Turing test next week but essentially it asks can you distinguish between a computer and a human.
As the internet grew in popularity, sites began to have a huge spam problem and a ‘Turing test’ had to be added to protect them from an onslaught of bots, by adding a human ‘check’.
Bots are just software running autonomously, often imitating human behavior but naturally working exponentially faster and more efficiently - hence why having a physical robot tick a box on a screen is not actually deceiving the system because you couldn’t achieve any spam-like effect that way, it would be incredibly inefficient.
You might remember that a few years ago the CAPTCHA looked like this?
But then bots and AI got too good for just distorted text…
So along came version 2, the mysterious check box, which as I said has little to do with the act of ticking the box.
Essentially it gets to work as soon as you are on the page/form, looking at things like how you got there, how you moved around, how quickly you entered information and proceeded, how quickly you moved the mouse and the directions of it, whether there are human-like cookies in your browser from other activity, and much more.
Exact details of how services such as Google’s ReCAPTCHA work are a secret, but the program comes to a conclusion as to whether or not you are a robot based on how you interact with the page.
If the system is unsure it will add a further test and this is where these images come in.
I personally use a browser with really strong privacy controls that block most cookies and all site tracking so I always, without fail, have to complete one of the fire hydrant and crosswalk exercises.
Here’s where things get really interesting though.
Some sites (whose owners have mal intentions elsewhere) might ask you to complete a CAPTCHA even though it knows you are a human, this is because, in reality, it is asking you to complete one on behalf of one of its bots… The CAPTCHA you might complete might actually have been sent from a bot that is on a different site and needs a human to do the last step, the answer is then sent back to the bot once you complete it 🤯
On a mass scale, this actually works well, it’s mad right?
The newest version, ReCAPTCHA V3, doesn’t show up at all but works entirely in the background and assigns a score to each user as to how likely they are to be human. This is the way forward because it further hides from bots as to how it is drawing conclusions, but also as A.I. and image recognition get better the CAPTCHAs will get too hard for us to complete, so it needs to be based on subtle actions.
Tom Scott made an excellent video that explains the CAPTCHA better than I can in a short few sentences here, so if you want to understand even more, take a watch.
I’m going to follow this all up next week, there’s so much I could add on the Turing test and the ‘intelligence’ of A.I. in 2021 that I simply cannot fit in this email.
AI can do some mad things now.
Saved 📷
Think 🌊
“Almost every car race movie has a side by side neck and neck scene where the protagonist proceeds to "floor it" to take the lead, as if that was not a suitable option before that moment?”
That’s all 🤙🏽
Feel free to reply to this email ↩️
I love to hear your thoughts, if you ever want to get in touch you can email guyb@cmeretome.ie or simply reply to any of these emails!
Please Share 📩🤝🏼
If you enjoy reading this every week I’m sure your friends would too.
Consider Supporting ☕
If you like my content and want to give back in the simplest way possible, please consider buying me a coffee.
Thank you so much ❤
Guy