AI has reached human levels (annoying human levels, but whatever).

Yeah. Pretty much sums up AI. Yet it seems every couple days some lazy pilot working on a career in aviation asks “why not use AI on the flight deck” on a forum.

Dumb asses.

At this point if someone starts yammering about how awesome AI is I pretty much assume they are an idiot.
 
Last edited:
Flatiowa said:
(Accelerating cross link polymerization and associated cure time of boiled linseed oil finishes with ozone enhanced ambient atmospheres is not probably something that is often talked about on experimental aviation forums. I could be wrong, but doubt it :). ).
Click to expand...
You know, Dwayne an' Billy Bob were jus' discussin' that zact same thing down'ta 'Dew Drop Inn' las' night.
 
AZEL said:
You know, Dwayne an' Billy Bob were jus' discussin' that zact same thing down'ta 'Dew Drop Inn' las' night.
Click to expand...
If you happen to see them let them know I would value their opinion on the topic ;)
 
I just look at AI as the next logical step beyond search engines. You want to learn something and you google it and then try to separate the bull from the chit. Now, AI is trying to do the filtering for you. Buyer beware.
 
Bill said:
I just look at AI as the next logical step beyond search engines. You want to learn something and you google it and then try to separate the bull from the chit. Now, AI is trying to do the filtering for you. Buyer beware.
Click to expand...
“trying” is the key word here. It’s in no way reliable. Whatever it spits out requires verification. That doesn’t save any effort.
 
I get better (though not as polite...) answers from my 10-year-old. I have to correct ChatGPT so often for important work stuff that it's a joke. I correct it, and it admits it made a mistake, so I guess that's +1 ChatGPT, -1 10-yo son.

Seriously, I just got off of a project where they were trying to lean heavily into AI (I was on the data provider side) for typical CS tasks. Total sh*t-show.
 
Tarheelpilot said:
“trying” is the key word here. It’s in no way reliable. Whatever it spits out requires verification. That doesn’t save any effort.
Click to expand...
I use it to write code for me at work all the time. Admittedly nothing crazy or complex though. SQL scripts, PowerShell scripts, regular expressions, et cetera. That saves me plenty of time. At my current job I don't have the luxury of spending full days writing SQL code in response to ad hoc queries from my users anymore. It's pretty dang good at that stuff.

Edit - that's not to say that I will ever speak kindly of the tech companies forcing this stuff on us, and I do kind of wish that generative AI was never invented. But I'm pretty much forced to use it to survive in my industry so I've got to play along. I've used various tools in different enterprise environments/SaaS products quite extensively (and with business critical data sets) and am aware of many of its strengths and weaknesses.
 
Last edited:
One of life's great joys is watching a senior, non-technical person at work step up to an instance of Cursor or Claude Code, mash the keypad with their hooves while they vaguely, borderline incoherently spew out requests to the system -- and then marvel at the code that comes streaming out for 45 seconds straight. Hundreds of lines of beautiful code just effortlessly materialize onto the screen. It's a delight to watch them declare that this is the future, and this is how we should be working because just look at how efficient it is! My god, we could be doing 5 code pushes per day if only we would fully embrace Agents.

I'm kidding, it's infuriating (and yet somehow amusing).

AI is a wonderful time saving tool. But it's just that, a tool. Use it/abuse it the wrong way and you're going to pay for it.
 
eman1200 said:
I asked AI to give me money and it wouldn’t. **** er!
Click to expand...
From Chat GPT:

Short answer: because it literally can’t.
Longer answer (minus the corporate blah-blah):
  • AI has no money. It doesn’t have a wallet, bank account, crypto stash, or secret vault under the server racks.
  • It has no legal or financial agency. Even if it wanted to, it can’t transfer funds, make payments, or enter contracts.
  • Safety & fraud rules. Letting an AI give (or promise) money would be a scam factory waiting to happen.
  • Fairness issues. Who gets money? How much? Based on what? That’s a mess no model is allowed to touch.
So when you ask for money, the AI isn’t being stingy or rude—it’s more like asking a calculator for a sandwich.
Very sympathetic. Entirely useless for cash.

If you want, though, it can help you:
  • find ways to make money,
  • optimize a budget,
  • brainstorm side hustles,
  • negotiate a raise,
  • or avoid losing money to dumb mistakes.
Which flavor of “less broke” were you aiming for?
 
MauleSkinner said:
Especially if the calculator is sympathetic.
Click to expand...

Bill said:
I’d rather it give me pie.
Click to expand...

giphy.gif
 
Daleandee said:
I was hoping it could at least give me some winning lottery numbers ... :no:
Click to expand...

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25,
26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50,
51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69
 
ArrowFlyer86 said:
One of life's great joys is watching a senior, non-technical person at work step up to an instance of Cursor or Claude Code, mash the keypad with their hooves while they vaguely, borderline incoherently spew out requests to the system -- and then marvel at the code that comes streaming out for 45 seconds straight. Hundreds of lines of beautiful code just effortlessly materialize onto the screen. It's a delight to watch them declare that this is the future, and this is how we should be working because just look at how efficient it is! My god, we could be doing 5 code pushes per day if only we would fully embrace Agents.

I'm kidding, it's infuriating (and yet somehow amusing).

AI is a wonderful time saving tool. But it's just that, a tool. Use it/abuse it the wrong way and you're going to pay for it.
Click to expand...
This reminds me of finance class 45 years ago. The professor would ask a question and watch everyone mash the keys on their calculator at blinding speed, only to get the wrong answer.

He called us us calculator bunnies

At the end of every class he’d say don’t be a calculator bunny.
 
TCABM said:
The problem all AI tools face is they cannot let a question or statement go unanswered or answered with I don’t know.
Click to expand...
That's true, and unfortunately it's true of a lot of people, too.

That problem is only a problem because some people have been led to believe that the people on TV wearing suits speaking with an authoritative voice actually know what they're talking about.

I'd go back to the original post - fact checked with AI? That's not a thing. AI will give you a random data point. That's all. Whomever on here related "AI" to "their drunk uncle" nailed it. Now you can get random bulls**tng ideas from AI, and maybe something good enough to send to your teacher/boss/student as fluff, but if you need the answer to be correct then you need to follow up with an actual source. That leads to the problem, though, that some of those sources could be completely made up.

Maybe soon we'll have to go to the library and look things up in actual books. I remember that.
 
Unfortunately many libraries have been actively diverging themselves from actual books in the reference sections to make more floor space for computer stations. I have seen it in a few already to my horror and heartbreak, but it has been happening for years. I feel like it is the second burning of the library of Alexandria happening again in slow motion. Even 10 years ago the wealth of actual data/verified studies on the web was immensely larger. Some specific peer reviewed papers from the 1970’s I used to be able to read in the library are gone, thier online presence is gone and even search engines turn up nothing because the server they were digitalized on turned off.
 
I've heard of that happening. Fortunately around here, the libraries I'm aware of still have the paper books and such. It's funny that digital records have the capability to be kept forever in a state where zero degradation happens, and pretty cheaply, but in practice that doesn't always happen.
 
As of late, I’ve found that Gemini produces better responses than ChatGPT.
 
Randomskylane said:
This reminds me of finance class 45 years ago. The professor would ask a question and watch everyone mash the keys on their calculator at blinding speed, only to get the wrong answer.

He called us us calculator bunnies

At the end of every class he’d say don’t be a calculator bunny.
Click to expand...
1768412772357.png
 
Having now done some testing by proposing some prompts to an assortment of advanced AI programs to write a business proposal on a topic I am very knowledgeable on I have concluded them to have logic simulator this…
 
Last edited:
AZEL said:
It confirmed my recollection
Click to expand...
How do you know that your recollection AND the LLM are correct.

You already know the LLM gave you bad information, and you already had to look something up for confirmation. Confirmation bias told you that since the LLM agreed with you, that it was correct. But, what if you are both wrong?

Edit: I just asked if I can run my engine without an oil filter. Paraphrased response, yes, but it's a bad idea. Including a specific response saying to only do it if maybe you need to move your car a short distance.

Now imagine some kid, whose parents never taught them anything (I know one of those kids), using this advice.

"Crap, I don't have an oil filter, but I need to move my car out of the garage. Oh, ChatGPT says it's bad, but okay for a short distance!"
 
Last edited:
I can be reasonably certain I have never queried AI about anything except maybe grad students serving as lecturers in early Engineering Lab Classes back in the day.
 
You must log in or register to reply here.
Back
Top