ClaudeCode, codex, opencode, etc. - these tools are about way more than just "code". Or equivalently, we vastly underestimate the versatility of code.
Just like mathematics explains so much of our universe - physicist Eugene Wigner called it "the unreasonable effectiveness of mathematics". Code plus the right LLM harness is unreasonably effective at knowledge work, but very few understand that currently.
Here's a random example. At a hunch, I threw in some raw astrophotography data (a FITS file) at VerbaGPT - and asked it to adjust the image to bring out the Orion nebula and the details. This is something that usually takes hours working with highly specialized software and requires a ton of image-processing skill. Neither of which I have.
VerbaGPT found an appropriate python package and applied dozens of edits (color stretching, contrast, complicated techniques like star-masking), and I think did a reasonable job! Took about 15 minutes, one of the outputs attached below.
What do you think? (about 10 minute exposure from my driveway, a Bortle 6 sky)
Originally posted on LinkedIn on January 15, 2026.