• TalkBass has been independent since 1998. Add your voice.
    Create a free account to reply to discussions, view embedded media, and browse with fewer display ads.
    Join freeLog in
    Want zero display ads or expanded classifieds tools? Compare plans.

I asked AI to design a bass cab

For fun, I asked Claude (4.5 Sonnett) your question:

Final Design Summary
ParameterSpec
Drivers2× Eminence CB3010N-8
WiringParallel = 4Ω
Power handling900W program
Box volume~26.6L gross internal
Tuning~66Hz
PortsTwo 2" diameter × 3" long
-3dB low~62Hz
High end3kHz+ (per driver spec, usable to 3.4kHz)
Sensitivity~97dB/1W/1m
Max SPL~120dB real world
Cabinet size17" W × 14" H × 11" D
Est. weight~28-30 lbs
Construction1/2" Baltic Birch, internal bracing

A 212 Neo cab, 28-30 lbs, hitting 120dB, tuned to 62Hz. Fits your amp, fits your five-string, fits in the back seat of most cars. The CB3010N-8 at 94dB sensitivity is actually slightly more efficient than most comparable drivers in this class, and two of them in a well-tuned small vented box is a very strong answer to your requirements.
The one caveat — at the box volume per driver I've calculated (~13L each), you're running on the lean side of optimal for this driver. WinISD would let you fine-tune whether a slightly larger box (say 30L total) gives meaningfully better low-end extension or whether the current numbers are the right tradeoff for the size constraint. Worth the 20 minutes to model it before cutting wood.
Would be interesting to get out pen & paper, design a cab using the same parameters, & compare it with the cab Claude designed.
 
Arent cabs all about getting the numbers "right"?
There are more parameters than just the numbers, for example there’s been no discussion of Q. Then there’s all the practical stuff…
 
For fun, I asked Claude (4.5 Sonnett) your question:

Final Design Summary
ParameterSpec
Drivers2× Eminence CB3010N-8
WiringParallel = 4Ω
Power handling900W program
Box volume~26.6L gross internal
Tuning~66Hz
PortsTwo 2" diameter × 3" long
-3dB low~62Hz
High end3kHz+ (per driver spec, usable to 3.4kHz)
Sensitivity~97dB/1W/1m
Max SPL~120dB real world
Cabinet size17" W × 14" H × 11" D
Est. weight~28-30 lbs
Construction1/2" Baltic Birch, internal bracing

A 212 Neo cab, 28-30 lbs, hitting 120dB, tuned to 62Hz. Fits your amp, fits your five-string, fits in the back seat of most cars. The CB3010N-8 at 94dB sensitivity is actually slightly more efficient than most comparable drivers in this class, and two of them in a well-tuned small vented box is a very strong answer to your requirements.
The one caveat — at the box volume per driver I've calculated (~13L each), you're running on the lean side of optimal for this driver. WinISD would let you fine-tune whether a slightly larger box (say 30L total) gives meaningfully better low-end extension or whether the current numbers are the right tradeoff for the size constraint. Worth the 20 minutes to model it before cutting wood.
With ports like that, it will chuff like a steam locomotive at about 100 watts. It’s basically unusable.
 
My comment is this: AI is a statistical classification and regression analysis tool, not an engineer. It is not capable of doing this kind of work, no matter how convincingly it fluffs you.
This.

LLM’s are prediction models coded to be inherently agreeable to the user: as a result, they routinely fail at even rudimentary problem solving tasks.

There’s a video currently going around of a guy asking ChatGPT how to use a water glass that’s upside down…and ChatGPT telling him repeatedly that a container which is open on the bottom and sealed on top cannot possibly be used as a water glass 🤣
 
Over time, I'm sure the AI's will improve, for now, trust but verify.

If you really want to build a Bass Cab:

Note, this is a lot harder than it looks, I know, I've done it.
 
Over time, I'm sure the AI's will improve, for now, trust but verify.

If you really want to build a Bass Cab:

Note, this is a lot harder than it looks, I know, I've done it.
It’s easy to design a crappy cabinet, it’s difficult to design a really good cabinet.
 
I’m more interested in what questions were asked to which AI, & what data sets were used by the AI. Could you include your questions that lead to your designs?
Can't speak for the op, but because of this thread I decided to take the AI plunge on this subject using Grok AI & Google Gemini AI.

For Grok all I asked was to design a front ported cabinet using two Eminence Deltalite 2515 speakers that will get me a sensitivity of 101db. Google Gemini I used the same requirements but added a LPad, crossover and Eminence APT-80 Tweeter. Both cab designs are tuned to 45hz. Grok used Python to run simulations for me as well. I didn't run any simulations in Gemini. Both used 3/4" Birch in their designs.

Grok designed a cab that was 36" tall x 24" wide x 18" deep, weight approx 75 - 85 lbs loaded (ouch)

Gemini designed a cab that was 38" tall x 21" wide x 15.75 deep, weight approx 72.3 lbs

That is external dimensions. Actual cut dimensions were also provided by both AIs as well.

For comparison, Mesa's Subway 215 is 34.37" tall x 19.25" wide x 19.5" deep and only weighs 60lbs

It was quite an interesting experiment to say the least. And I got a lot of data out of this. Now we just need a cab designer to look it over to see how accurate it is. But we don't have to worry about me building one, too heavy. And that doesn't include the added weight of handles, a grill, or something like duratex paint. Hey @agedhorse wanna take a look see, like we don't bother you enough as it is already 🤣

I'm going to post all the pages of data as thumbnails, Grok on a black background, 14 pages. And Gemini is on a white background, 10 pages. You'll definitely need to be on a decent monitor to scroll through these.


Grok(In Expert Mode):

001.jpg
002.jpg
003.jpg
004.jpg
005.jpg
006.jpg
007.jpg


008.jpg
009.jpg
0010.jpg
0011.jpg
0012.jpg
0013.jpg
0014.jpg



Gemini:

001.jpg
002.jpg
003.jpg
004.jpg
005.jpg
006.jpg
007.jpg
008.jpg


009.jpg
0010.jpg
 
Last edited:
Wouldnt surprise me one bit if AI was already involved in some form in Cab and Amp design.

---

Social media's idea of what AI is and what its capable of is likely a bit skewed by the very nature of Social Media where the user is the end product.
People often want to believe there's something essentially different about human thought, maybe consciousness, qualia, intentionality, or free will.
But "it's just pattern recognition" doesn't identify that difference, because it applies equally well (or equally poorly) to both humans and AI.
 
  • Like
Reactions: stereobate
Can't speak for the op, but because of this thread I decided to take the AI plunge on this subject using Grok AI & Google Gemini AI.

This thread was a cautionary tale not to waste time and resources feeding this AI delusion. You just raised the electricity bill for a bunch of people in rural towns already on fixed incomes.
 

Latest posts