See
The code is in LOGO, and draws a picture.
…and see this project on GitHub where they have run the first 140min of code from the audio… so far…
This what it looks like currently:
Yeah, I listened to that for a minute or so, closed the page, and came to look over here.
LOGO! Elementary school flashbacks.
Wow! Turtle graphics!
Here’s the final picture:
The audio file is 9 hours and 7 minutes long. 
Turtles all the way down… 
However, the code isn’t. It styles itself as if it is a TV show telling you how to paint the scene yourself. Just scrubbing through the audio file, I found a lot of asides with various facts and stuff. It’s supposed to be an electronic Bob Ross show.
Thanks for posting. I would’ve just assumed it was another broken xkcd comic. For some reason, whenever the author tries something fancier than a simple image, my browser can’t handle it. (Probably due to firewalls.)
My word is “usufruct”…
Is it yours, or are you just using it?
It’s mine … my own… my precious…
Can I use it? I promise to treat it well and wisely.
I think your right to use it may be inherent? I can’t be sure though. I couldn’t find a coherent antonym for “usufruct” other than unusufruct
And it now appears in a thread quote in the There’s been a Communist revolution and your house now belongs to the people.
See a quote from md-2000, post 10 from a google article. “…obtained the right of perpetual usufruct of the land, plus the ownership of their homes…”
I guess I get to do this one…
To be fair, both Victor and Adam would have the last name “Frankenstein”.
But if there’s a reason for the other 12 panels, I’m not sure of the joke.
It’s about pedantry.
From explain xkcd:
The Captcha shown in the comic instructs the user to select all tiles containing Frankenstein. The tiles include both a reanimated corpse resembling Frankenstein’s monster and a scientist yelling, "It’s alive!” who is clearly intended to be Victor Frankenstein. The problem arises from the contrast between the generally accepted and technically accurate definitions of the term Frankenstein. The correct answer to the Captcha is just the left square of the third row, unless you follow comic 1589. If the images in the squares are scenes from the famous 1931 film starring Boris Karloff as The Monster, then they could be correctly said to be "containing Frankenstein ”—that is, the work. Because of the ambiguity regarding what Frankenstein refers to, this would not be a good CAPTCHA because many people solving it would use an incorrect definition of Frankenstein and therefore get it wrong. (However, it would be effective in screening for people who know that Frankenstein technically refers to the scientist, not the monster—or, if one also had to mark the boxes depicted Frankenstein’s monster, screening for people who don’t know.)
Many of the other tiles appear to be pictures of entities that inspire similar pedantry. For example, there is a picture of a turtle (or possibly a tortoise, or a reference to the Voight-Kampff test used in a manner analogous to CAPTCHA), a ship (or possibly a boat), Link (the name given to each of several protagonists that appear across generations and timelines, throughout the Legend of Zelda video games, who many erroneously refer to as Zelda), a pond (or possibly a lake, or a mirage), a squash or pumpkin (often subject to the fruit or vegetable debate), an erupting volcano (with lava, or is it magma?), and an asteroid or planet (or is it a dwarf planet?). Other tiles seem to be inspired by images that commonly occur in actual captchas, like the STOP sign or the traffic light. However, at least some of these may also be meant to fall into the category of entities that inspire pedantry, for exmaple: because traffic lights can also be called traffic signals or stoplights; many people thinking that the shape of a stop sign is a hexagon, not an octagon; and the definition of a sandwich (previously discussed as a “random semi-ironic obsession” in 1835).
Of course, a real pedant knows that none of the squares contain Frankenstein, since each square is only a picture, and a picture is not a scientist (or animated corpse, etc).
That’s a pretty subtle joke.
Those captchas must get a huge variety of responses anyway. What counts as selecting a streetlight? Selecting just one block that contains most of the light assembly? Also selecting the block that contains 3 pixels of the light assembly in one corner? Do you select all the blocks that have the street light pole within them? etc…