You must log in or register to comment.
I know we’re all AI fatigued, but this seems to be the good, old school ML type,
How this AI was trained
Venture Beat’s coverage contrasts Quilter AI against LLMs like GPT-5 and Claude. Indeed, circuit board design isn’t a language task or problem. Thus, the AI behind this tool is basically trained by playing an optimization game against the laws of physics.
Surprisingly, there were no earlier stages where Quilter AI was trained on human-designed sample boards. This decision was made because humans frequently make mistakes in board design, and to make sure Quilter AI’s capabilities weren’t somehow capped at human-level.
Massively reducing invested human time to solve logical problems. Again, according to the article, down from 430 hours over 3 month development to 38.5 human-assisted hours over a week.
I am sure that by AI here they mean not LLM but a large special model that generates boards designs, not words, so we can’t really judge this type of model based on LLMs.
deleted by creator
I can believe this, I can’t get standard LLM models to design a board worth a damn.
You can’t, really?
Did you actually try to get to a Gerber from there? I’ve done this a few times with Kicad_sch and it just doesn’t get to anything useful.
It’ll tell you everything you want to hear, but try to use it to finish the job. I’ve tried to get it to just wire up a some simple MCU/sensor schematics to import and it’s a mess.
