Artificial Intelligence

nola mike

Vice Admiral
Joined
Apr 22, 2009
Messages
5,980
Use AND, NOT, and other boolean logic and the searches do not respond as they used to.
I'm not familiar enough probably with anything complex but Google apparently does support at least some operators... What can you not do now that you could previously?
 

Scott Danforth

Grumpy Vintage Moderator still playing with boats
Staff member
Joined
Jul 23, 2011
Messages
52,819
I'm not familiar enough probably with anything complex but Google apparently does support at least some operators... What can you not do now that you could previously?
one example

"C4" AND "Dana" AND "36" AND "ring and pinion" NOT "ford"

and you get all ford, and none for a Dana 36

changing to +/- and you get a different result where the result should be the same
 

nola mike

Vice Admiral
Joined
Apr 22, 2009
Messages
5,980
one example

"C4" AND "Dana" AND "36" AND "ring and pinion" NOT "ford"

and you get all ford, and none for a Dana 36

changing to +/- and you get a different result where the result should be the same
Yes, but you can use +/- and get the desired behavior, right? Just can't use actual boolean operators?
 

Scott Danforth

Grumpy Vintage Moderator still playing with boats
Staff member
Joined
Jul 23, 2011
Messages
52,819
using the + / - doesnt give the results I am looking for. just different than the boolean operators.
 

BWR1953

Admiral
Joined
Jan 23, 2009
Messages
6,583
It.
Has.
Begun!

I only use the basic AI that comes with Google Chrome. It's the non login version of Gemini. I have many threads with a wide variety of topics. It has never been able to "see" the other threads before.

This evening I was talking about one thread, got some answers and closed that tab. Then I closed all my other tabs, plus the window and opened a fresh window. I then typed a single word to use as a reference and hit enter. And I got an immediate response which included my conversation from the prior thread! And that response was related to my prompt word.

I asked it how it got that information because it had told me previously that it didn't have that capability. The answer was a list of recent upgrades showing that it now has memory of and access to ALL of my current and prior threads! 😲

It's started.
 

Scott Danforth

Grumpy Vintage Moderator still playing with boats
Staff member
Joined
Jul 23, 2011
Messages
52,819
We just banned an AI chatbot that was simultaneously posting on iBoats and at least 7 other forums

Come on solar flare and big EMP
 

nola mike

Vice Admiral
Joined
Apr 22, 2009
Messages
5,980
Interesting. Thought there were AI responses and bad information, but didn't think it was a bot.
 

redneck joe

Supreme Mariner
Joined
Mar 18, 2009
Messages
15,593
so, what would be the purpose of a bot, here on this site, be?

We have no social info, no financial, not monetized, etc?
 

nola mike

Vice Admiral
Joined
Apr 22, 2009
Messages
5,980
ok, so real world testing incoming...this is a lot of text, but whatever.
My 25 hp merc died suddenly (again). 2 years ago had similar symptoms, turned out to be the stator. Information provided to chatgpt:

1. Weak spark to both cylinders, unable to jump gap on tester
2. DVA voltage on both stator leads 180v while hand cranking (barely within spec, but presumably very low rpm). Resistance tests normal. (I was guessing stator since chinesium replacement and same symptoms, but seems ok. )
3. All connections checked, grounds cleaned
4. Trigger resistance 750k ohms, spec 6.5-8k ohm.

Suggestion:
"Since the spark is equally weak on both cylinders, the trigger is not the leading suspect."
"Most likely order in your case: switch-box or coil grounding → partially grounded kill circuit → switch box itself → two degraded coils/plug wires → trigger."



I had already mentioned kill circuit was disconnected.
It kept suggesting things and I would offer reasons why I didn't agree. It eventually turned 180':
"
790 kΩ is not just somewhat high—it is effectively an open trigger winding if you measured directly between the trigger leads with the trigger disconnected from the switch box."

"A normal reading for this style of 25 hp Mercury trigger is generally in the hundreds of ohms, roughly around 650–850 Ω, not 790,000 Ω."

ME: "specs say trigger resistance is 6.5 k ohm, not 650 ohm. reading is trigger disconnected from switch box"

"
You’re right—I gave you the wrong resistance range for this ignition.


With a specified trigger resistance of about 6.5 kΩ, a measured 790 kΩ with the trigger disconnected is grossly out of specification—roughly 120 times too high. That indicates an almost-open trigger winding or a failing connection within the trigger leads.


That makes the trigger the strongest suspect, especially given:


  • Sudden onset
  • Weak spark on both cylinders
  • Approximately 180 V DVA from the stator while connected
  • No change with the stop circuit disconnected
  • Both cylinders sharing the same trigger assembly"
"I would not spend time blaming both coils or the plug wires first. The coil-input DVA test would still be useful, but that trigger resistance is already a clear failure finding."


Now, I actually don't know what the problem is at this point. I ordered a new trigger assembly. However, had I just listened to AI I would not have really suspected it. I had to correct it multiple times (both reasoning and data), and then it agreed with me. I actually don't feel great about the trigger as the cause, but my diagnostics say otherwise. As a side note, this motor was running absolutely perfectly for the first time since I've owned it for exactly 2 weeks. Curious what the actual problem is, we'll see (hopefully).
 
Last edited:
Top