Love this idea!

As someone with a couple of code bits tattooed on my body, if anyone is planning on more than a few lines of text, speak to your tattoo artist first.

I once wanted a long bit of code on my arm. I thought it was a good idea, but speaking to my tattoo artist, he explained that after a few years, the text would likely blur and merge together until it was difficult to read.

Ended up going with a smaller snippet which is still clear to this day (15 years later). And seen some others who didn’t listen to advice with an illegible blob where their clean code one was xD

• cinntaile

  today at 6:33 AM

  They should have read Uncle Bob's book.

Definitely Andrej Kapathy's 250 line code of Micro GPT.

I think it can fit on someone's back. (if a small enough font is used)

- https://gist.github.com/karpathy/8627fe009c40f57531cb1836010...

- https://karpathy.github.io/2026/02/12/microgpt/

• goodmythical

  yesterday at 10:21 PM

  My brief search says nasa says the average back is 47.6cm * 39.2cm. ~18.74in*15.43in is 289.152in^2.

  Code in question is 9319 characters counting whitespaces according to copy/paste github raw to zed and selecting all characters.

  That's ~32.2 characters per square in.

  Aside from the likelihood that it smears/fades to illegibility in a few years, bless the artist that has the patience.

  Hell, good luck finding anyone, really. Any artist can throw some words on you, but to do this scale... Getting every line to be "straight" across the back over 200 lines, with each character taking perhaps 15 seconds (because this would require precision single needle work with a good steady hand), that's just shy of 40 hours, which I suppose is in line with typical times for large detailed pieces.

  Anyway, I guess what I'm saying is go look at some declaration of independence tattoos because that's ~17% as many characters (1628 without the signatures) meaning MicroGPT as a tattoo requires nearly six times as many characters per inch of skin.

• lawlorino

  last Sunday at 1:47 PM

  That’s lines of code, not weights

Don't do this, it's equivalent to that barcode tattoo that passed it's heyday.

I don’t think you can get a more simple ML model than linear regression?

y=w0+x*w1

Encode it into a barcode