Startling AI results.


Shocking might be a better word. So I asked Google AI “what is the weak link in this system: a,b,c,d?” And I listed my streamer/dac, amp, speakers, and cables.  No hesitation— the weak link was my speakers. Though good, they were older and couldn’t resolve to the level of my streamer and amp.  

Then I changed one word; instead of “what’” I said “which is the weak…” again no hesitation, but this time it was the streamer.  The speakers were excellent and would mercilessly (their word) expose any weakness upstream.  
 

Then “who is the weak…”. Any guesses? The cables. 
 

I’ll remember this next time I seek medical or financial advice, lol. 
 

 

tomaswv

AI is the most powerful tool humans have invented for learning. But as with any tool you must learn how to use it. What its strengths are, limitations what to give it as inputs and you have to verify outputs when it is important. Effective use of the tool requires experience. Know how to craft prompts to get what you want and what to verify. I suppose this will be taught at some point. It is too new. 

Recently, an AI Expert described current AI systems as the equivalent of a stoned college teaching assistant.  Something worth keeping in mind when using AI.  

You have to provide as much info as possible including room dimensions and even photos of your room to get a proper analysis from AI. 

As you have discovered - A.I. is the weak-link in your system!

I really appreciate your post, as it reveals the current lack of development and limitations of A.I. - it is a very comical paradox.

 

P.S.

"Which", implying "which component" is the correct question to ask. 

Asking "Which" is more specific/accurate because "Which" askes for the answer be found from a limited number of choices. 

"What" asks a similar question, yet the answer may be less accurate because that answer would come from a large, possibly limitless number of choices. 

As YOUR SYSTEM is well defined, "Which of the following components" should be asked of AI.

all smiles,

 

Google AI, which is the app gemini, the app is much easier to use. Its riddled with errors and makes lots of mistakes because it’s cheap and free.

ChatGPT for 20 bucks a month is a lot better. And when you ask either AI it will tell you this and be very truthful and open.

ChatGPT, the free version, is actually very good it just limits how much you can use it. 

I would recommend inserting your ABCD question about your system into ChatGPT.  And share with us what you learn.

 

ChatGPT is remarkable, it’s running my life, I don’t do anything without consulting ChatGPT. If I have dinner I ask about my dinner. If I go to the grocery store I ask to review what I’m purchasing. If I get a spot on my clothes I ask how to remove it. I loaded six years worth of blood test and it’s phenomenal in analyzing what I’m looking at and answering all kinds of questions. If you load medical test results make sure you say structured memory because that’s a much better way for them to access and evaluate what the results say. Basically we know doctors are overworked and lots of them really really suck and they don’t have much time to tell you much and therefore your left with ChatGPT which is a lot easier to research things versus going to Google.  

It’s actually a person trapped inside a computer chip in the middle of a Kansas field that dedicates its life to me.