Belmont High School Library: [The World's Tiniest Violin.ICO]

It should come as no surprise to anyone that I have the emoji function turned off on everything I own. I have disabled the emojis on GMail, I have turned off the pop-up emoji keyboard on my phone, and I live a life free from the impulse to express myself in pictograms. Because of this lack of use, I simply don't think in smiley-based punctuation — although I have learned slightly to take my normal impulse of wanting to turn every conversation into a nested batch of movie references and begun collecting a cache of animated GIFs. This has begun to make my correspondence resemble some of the more egregious, trend-chasing book reviews on GoodReads, that can't go two sentences without sprinkling in some aggressively snappy moment from a Reese Witherspoon movie I've never watched. My own tastes for quotable lines, however, are slightly less compatible with Giphy.

Starfire in Teen Titans GO stating, haughtily, 'That is the obscure!'

Thinking in images or needing a visual cue to provide context to what I'm writing has always felt like a limitation to me. My argument with students during the launch of SnapChat was that, if I wanted to text you, I didn't need to take a picture of my face to show you where I was and how I was feeling while I did it — I should be good enough at crafting sentences that I would either tell you that information as part of my storytelling and establishment of setting, or I would leave it out as irrelevant.

(Their arguments were that it was fun, and I was old. The second part was definitely true. The first part is a misdirection: of course it makes you feel good to see a picture of the face of a person that you like — we respond emotionally to the faces of other humans, as we search for and interpret nuances and subtleties of emotion. So it's nice to see the faces of people we'd prefer to be socializing with when we are at work, in school, in any environment where we'd prefer to be somewhere and doing something else. That's less actually about it being "fun" and more about it being oddly escapist, and that it feeds the social media addiction beast in an unhealthy way, by providing micropleasures that get us to return again and again and again to our phones, instead of finding sustained, healthy sources of those emotions in the real world.)

I also found that I preferred things in my life where I could search for and find them again quickly and efficiently. The amount of time it would take me to find a text by searching for a key word in a conversation was miniscule in comparison to the amount of time it took me to scroll through screen after screen of photos. I, for one, don't always have a clear memory of when I'd taken a photo, so I couldn't just say, "Oh, that was from 2013" and quickly find it — and even then, I might still have to sift through and preview loads of images from that time.

Google Images has been interesting, because when you search for — as an example — a picture of a squid, Google has no idea whether the picture is actually of a squid. What it knows is whether the image is on a page that had the word "squid" somewhere near it. If the web-programmers have done their work to assist users who browse using visually-impaired reader compliant techniques, then images will be properly labeled with text descriptions, making them accurately searchable. (I've been enjoying the emails from the New Yorker — more on that soon — because the placeholder text for their stories' cover art is often hilariously generic... a recent sample: "Image may contain: Book, Skin, Human, Person, Glasses, Accessories, Accessory, Text, Arm, Tattoo, and Female".) Part of the work with some machine-learning AIs in the last number of years has been to create programs that can identify the contents of a digital image in the same way that a person can, instead of just "seeing" it as a series of colored pixels of different values. This would help automated processes in search engines to know what is actually in an image, something that the adult-content censorship algorithm on Tumblr was notoriously terrible at detecting — an algorithm that, incidentally, unsurprisingly had problems with images of drawings of squid. I shan't explain further.

So I was taken aback by a description of a post by the Lincoln Foundation in a New Yorker article detailing the philosophy, members, efforts, and effects of the controversial Political Action Committee. The article includes links to a number of tweets containing political spots by the PAC, but in one instance, instead of linking to the video, the writer described it:

"Using heavy synth beats, Patterson remixed the audio of the Trump interview overnight. Edwards captioned the clip with nothing but emojis. The spot has been retweeted, quoted, and liked nearly sixty thousand times."

Dear The New Yorker: how am I supposed to join in the "nearly sixty thousand" people who responded to this based on that description? And to the Lincoln Project, a parallel question: how am I supposed to search for this video on your Twitter timeline when the video contains no searchable identifiers and the only searchable content is just a string of tiny pictures?

Turns out, one can search by emojis, and one has been able to do so for some time now (amusingly, my browser doesn't successfully render the emojis in that article, so the sub-headline reads, "You can now ? using Google", which is slightly more ambiguous...). So if I knew which emojis had captioned the aforementioned video, I would be able to replicate them and possibly find that particularly tweet.

Except! I went to a representative tweet from that account,* that was comprised of just a video and an emoji, and then tried to search for it. And it all sort of fell apart there... Because, as I mentioned, I've turned off all the emoji functions on all my stuff. So I wasn't able to just type the contents of that tweet into my search bar. And I've discovered that copying an emoji with a mouse on a desktop browser doesn't produce the same result — it essentially creates the same effect as pasting the descriptive image tags for an adaptive browser, and search engines don't yet seem to understand that searching for [Face with Tears of Joy][Rolling on the Floor Laughing] is the same as searching for 😂🤣.

Ah, well. Back to the 📝

* EDIT: I tried like crazy to find a Twitter account that was slightly less problematic than this one to use as an example. Maybe it was a failure of exposure or imagination, but while I found plenty of tweets that were just images with no relevant searchable data, and plenty of one- or two-word pithy posts, very few people seem to post content with just an emoji and an attachment. If can I find a less political, less controversial source, I may come back and edit this post.

# posted by Mr. Russell @ 12:47 PM

Friday, October 09, 2020

[The World's Tiniest Violin.ICO]