lb3 said:

---

Your right. I'm no code monkey and haven't written any notable scripts since 2016 when my excel VBA script that archives TexAgs history made the admins introduce captcha on frequent edits.

---

Is there a way to capture or save an entire thread? Some are spectacularly informative and that would be good.

My script just captured all of an individual user's posts and allowed them to be mass edited or restored if needed. It's a bit clunky to run since it requires internet explorer and a registry edit. DM me if you want a copy.

TexAgs91 said:

---

Just wait until the humanoid robot revolution starts

---

I don't think it will completely take over.
At some point customers will push back and not visit places where they are served by a robot. Those will go out of business and we will come to a reasonable middle ground.

lb3 said:

---

Your right. I'm no code monkey and haven't written any notable scripts since 2016 when my excel VBA script that archives TexAgs history made the admins introduce captcha on frequent edits.

---

I noticed the annoying captcha yesterday when I had to make edits.
They bug me like hell.

We are redoing our IT platform right now using AI. Generating code in at least a quarter of the time with far fewer people than we otherwise would have hired.

It's a game changer for sure.

I work in commercial strategy and I use it as a strategic partner to get from 0 to 1 faster, not all the way to 100.

Like plenty others have said, AI will not take your job. Someone who knows how to use AI will.

infinity ag said:

---

lb3 said:

---

Your right. I'm no code monkey and haven't written any notable scripts since 2016 when my excel VBA script that archives TexAgs history made the admins introduce captcha on frequent edits.

---

I noticed the annoying captcha yesterday when I had to make edits.
They bug me like hell.

---

I'm sorry about that. The admins don't want people being able to replace hundreds of old posts with advertisements or porn.

AI is a tool and pretty nifty one.

I'm going to guess it didn't outright fabricate but rather found someone else's lie it dug up convenient to its purpose and ran with it. I am going to guess most AI don't attempt to sanity check or cross reference and validate source material for plausibility. That would be an very advanced reasoning task.

I disagree only because I've seen it out right make **** up. That's the biggest problem I have, it just doesn't know how to say "I don't know".

That's one of the issues, AI has the confidence of a drunk guy at a bar who wants to tell you about his business idea. You have to put in some work to suss out the BS.

MouthBQ98 said:

---

I'm going to guess it didn't outright fabricate but rather found someone else's lie it dug up convenient to its purpose and ran with it. I am going to guess most AI don't attempt to sanity check or cross reference and validate source material for plausibility. That would be an very advanced reasoning task.

---

It's called a "hallucination" in the parlance of the AI people...

The interesting thing is that it was "partly right".

The document numbers on both were ALMOST correct. The last 2 digits represent the year the latest revision of them came out.

No revisions came out those years.

And the subject of the documents was "somewhat" related...


My point was to make sure you double check EVERYTHING. Because if it can make THAT hallucination, it could make a worse one....

bthotugigem05 said:

---

That's one of the issues, AI has the confidence of a drunk guy at a bar who wants to tell you about his business idea. You have to put in some work to suss out the BS.

---

Wait...

So you're saying that ChatGPT isn't the first Green Beret SEAL?