AI learns the “dark art” of RFIC design • Hacker News Board

Re: AI learns the “dark art” of RFIC design

#191

Post by nz » Wed, Jul 01, 2026, 12:07 AM UTC

deterministic wrote: post #189

nz wrote: post #120

Earlier quoted context omitted.

As someone who does a lot of genetic programming (like, old-school, without AI/LLMs, etc), I can confirm that the fitness function is very difficult to get right, especially if you are trying to evolve programs that have "adversarial fitness" -- you'd need to maintain a hall-of-fame, and that just makes the runs take _much_ longer, because, chances are, your fitness function is the bottle-neck. So, it is very hands-o…

Maybe you could have generated random test input, generated the trees, and then converted the trees back to the input, checking if they match?

Yeah, that is situation 2 mentioned above. A Chomsky Grammar is also a generator. So it can generate valid inputs, and then turn them into valid outputs -- and it can do the first part stochastically/randomly.

Re: AI learns the “dark art” of RFIC design

#192

Post by jcims » Wed, Jul 01, 2026, 5:42 PM UTC

gooseyard wrote: post #108

jcims wrote: post #18

Reminds me of this old article - https://www.damninteresting.com/on-the-origin-of-circuits/ One of my favorite little morsels of internet goodness.

I was going to post this as well, its delightful to see that other people enjoy it since it was really mind-blowing when I read it. It's interesting since I saw another comment near yours that raised the question of robustness of the lab-grown design, which I thought was kind of the most fascinating part of the damninteresting article was the revelation that the evolved programs were inseparable from the single physi…

Great comment!

>I thought was kind of the most fascinating part of the damninteresting article was the revelation that the evolved programs were inseparable from the single physical FPGA used in the training.

100% agree with this!!! It gave me this weird feeling the first time i read it, like the onset of some alien intelligence. xD

I definitely think a simulator is the way to go, but I'm guessing tools like that could find problems and edge cases in the simulator that nobody thought to test for.

I'm just glad they still have the article up. I bet I've shared it 50 times over the past 20 years lol.

Re: AI learns the “dark art” of RFIC design

#193

Post by gamblor956 » Thu, Jul 02, 2026, 6:21 AM UTC

amelius wrote: post #164

gamblor956 wrote: post #134

Earlier quoted context omitted.

Wouldn't work. Judges would not treat the AI generated designs as prior art without proof of human involvement (above and beyond entering the prompt).

Maybe not prior art but very much "not novel".

If a human was not involved then it remains novel to humans.

Otherwise someone could copyright every combination of words.

Re: AI learns the “dark art” of RFIC design

#194

Post by Brian_K_White » Fri, Jul 03, 2026, 7:52 PM UTC

xp84 wrote: post #190

Brian_K_White wrote: post #167

Earlier quoted context omitted.

Utterly haphazardly and inconsistently of course, same as yours. You thought that was some sort of argument? It slots right in and contradicts nothing.

I mean you can't explain how your consciousness works, so by your logic you don't "know that it works." But you do know that it works, because here you are.

In fact I do not know that it works because we don't even have a definition for works for that. All we can say, maybe, is that it exists. How well it works is up for grabs but my own strongest suspicion is "with a great many flaws", which I don't consider "works" in most other contexts.

The comment was intented to be pithy, which means OBVIOUSLY, it has to generalize and be more poem than encyclopedia. A few words that express the essense of an idea. That is both useful and true even if such an essesnse OBVIOUSLY leaves contrived edge case exceptions unadressed. Congradulations on spotting that the universe is actually infinite.