Tunc Drive (Python DSP - via GPT)

This started with a question I asked ChatGPT out of curiosity: could a simple prompt get me a playable overdrive in a .nam file that I could load on my HeadRush Prime? I tried the first file, played it on the Prime, and went back with two rounds of feedback. I’ve included all three versions so you can hear what changed. 01 — Prototype Drive. It worked, which surprised me. I had to turn up the input quite a bit before it started to break up. 02 — High Gain Drive. I asked for more gain and for the breakup to respond to how hard I picked. It was easier to push, but the output was too loud in my rig. 03 — VH Drive. I asked for more drive, lower output, more body around 500 Hz, and a small lift around 2.4 kHz. The filename reflects the gain target I gave it. But it doesn't really come close to a VH drive in terms of gain. In my rig, it landed closer to SD-1 territory. I’m happy with it as a blues overdrive or as a push in front of an amp that’s already breaking up. If I hit the input very hard, it starts to get a little fuzzy, and the attack softens in a way that reminds me of tube amp sag, which is what I was going for. With my prompts, ChatGPT wrote the Python code and calculated the NAM A2 weights directly. The design uses input gain, progressive asymmetric clipping, and output filtering and EQ. There was no physical pedal, reference recording, or NAM training involved. My part was to play each version and describe what I heard and felt. The models are cabinet-free. I use them before the amp block on my Prime. I don’t support using AI to create art. I tried this because I was curious about what it could do as a tool for building a guitar effect after watching a Rhett Shull YouTube video. To me, these files don’t have the same authenticity as a carefully made capture of real gear. There’s a craft to choosing the gear, dialing it in, and capturing how it responds. But I’m still happy with how this experiment turned out, and I thought it was worth sharing all three versions so you could hear it for yourself. What are your thoughts on this? I'd love to hear your feedback.