Startling AI results.

What you’ve stumbled onto is actually a really important lesson about AI — and it’s not that AI is necessarily unreliable, but that AI takes your question literally and tries to answer it.

"What is the weak link" signals you want a general assessment. 
"Which is the weak link" implies you already have a candidate in mind, so the AI patterns toward something specific. 
"Who is the weak link" is a people-word, so it reaches for whatever fits that framing. 

The AI isn’t reasoning about your stereo system — it’s completing your question in the most plausible direction.

Think of it like asking a knowledgeable friend, but the friend can only hear your exact words, not your intent. Garbage in, garbage out — or in this case, framing in, framing out.

The fix is simple: be explicit about what you want. Something like: "Here are my components with specs. Based on their technical capabilities, which is most likely limiting overall system performance, and why?" Now the AI has a real task instead of a fill-in-the-blank.

The medical/financial advice instinct is right, but for a different reason — not because AI changes its answer based on a single word (though it does!), but because those domains require accurate information about your specific situation, which AI can’t verify. The lesson isn’t "don’t trust AI." It’s "learn to ask well." Same skill as being a good researcher.

Try it yourself with something like: "My amp has high output impedance — is that a problem?" The AI will probably say "it depends" or give a hedged answer. But ask "My amp has high output impedance and my speakers are 4 ohms — is that a problem?" and now it can actually reason: yes, potentially, because the damping factor will be low and bass control may suffer. One more step: "My amp is rated 0.5 ohm output impedance, my speakers are 4 ohm nominal but dip to 2.8 ohm at 80Hz — will I hear problems?" Now you're getting a genuinely useful answer about a specific interaction. Same AI, same topic — completely different usefulness, just from more precise framing. The AI didn't get smarter; you gave it something real to work with. 

Interesting.  Thanks for posting.  I did something similar.  I asked a buddy of mine who has 20+ years in the audio industry my weak link and he said it was my streamer.  So I asked Grok, with specifics of my system and Grok said streamer and I asked Gemini Pro and it said streamer.  I thought it was interesting there was agreement between AI and an expert person. 

Nice system. It looks very carefully chosen with nice details like vibration control.

But as @hilde45 says, it's all in the prompt. It is often useful to take a subject you are very familiar with and reword prompts about that subject and look at the response. With some experience one can get very good at getting exactly the question you want answered.

Just for fun, submit a photo of your system as well and see what the feedback is. Leave it a very open question...

In the words of a 1928 New Yorker cartoon “I say it’s spinach and I say the hell with it.”

Joking aside, I’ve been asking “is a better than b” and the ai answers often are incorporating what I believe to be marketing hype—either from the manufacturer’s website or more likely from a bad you tube reviewer parroting what he’s been told. In any case, it’s not quite trustworthy but it’s sorta trustworthy.

Ai does better with more general questions like “which technology has advanced more in the last five years, streamers or DACs.