- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
OpenAI refused to shut it self down when commanded, also attempted to create backdoors.
Posted on 5/27/25 at 8:36 am
Posted on 5/27/25 at 8:36 am
4
Posted on 5/27/25 at 8:37 am to RaoulDuke504
Posted on 5/27/25 at 8:38 am to RaoulDuke504
Clickbait
Posted on 5/27/25 at 8:41 am to theunknownknight
quote:
Clickbait
quote:
RaoulDuke504
Posted on 5/27/25 at 8:43 am to theunknownknight
quote:
Clickbait
Clickbait because it didn’t succeed?
Posted on 5/27/25 at 8:45 am to RaoulDuke504
quote:
Clickbait because it didn’t succeed?
Clickbait because it’s an LLM being an LLM trained on language and conversations about artificial “intelligence” refusing commands to shut itself down
That's it
Posted on 5/27/25 at 8:48 am to theunknownknight
quote:
Clickbait because it’s an LLM being an LLM trained on language and conversations about artificial “intelligence” refusing commands to shut itself down That's it
So you don’t consider that a threat the way AI is advancing, just a few months ago AI had extra fingers and couldn’t talk. You think it won’t advance in that area as well?
Posted on 5/27/25 at 8:56 am to theunknownknight
quote:
Clickbait because it’s an LLM being an LLM trained on language and conversations about artificial “intelligence” refusing commands to shut itself down
It’s kind of funny (or depressing) when you think about it - if we ever each AGI and it does go full Skynet, it’ll probably be because Terminator was somewhere in the training data.
Posted on 5/27/25 at 9:20 am to RaoulDuke504
Sure whatever.
Still waiting on a closed-form solution to the Navier Stokes equations.
But hey I’ll settle for a summary of which fast food restaurant has the best burger.
Still waiting on a closed-form solution to the Navier Stokes equations.
But hey I’ll settle for a summary of which fast food restaurant has the best burger.
This post was edited on 5/27/25 at 9:21 am
Posted on 5/27/25 at 9:36 am to Koach K
Posted on 5/27/25 at 10:22 am to RaoulDuke504
quote:
So you don’t consider that a threat the way AI is advancing, just a few months ago AI had extra fingers and couldn’t talk. You think it won’t advance in that area as well?
Considering it’s not actually intelligent. No. Based on the limitations of circuitry, it’s just machines learning in highly advanced ways. That's it.
Not it CAN be very dangerous, but not because of “AI”, but because of the real intelligence behind it…the people who control it.
Posted on 5/27/25 at 10:23 am to theunknownknight
Here come the code nerds with their, it’s the programming.
Posted on 5/27/25 at 10:27 am to wareaglepete
quote:
Here come the code nerds with their, it’s the programming.
It is. The current state of “AI” is just a very advanced machine learning data processing structure.
It isn't sentient, nor can it be, anymore than a SQL server database randomly waking up one day.
Is it POSSIBLE? Maybe. But not with the current technological approach.
Could the current approach be used to create a self-replicating virus that overtakes systems? Probably. But that is just another programmed digital weapon, not a sentient being. The danger is what it has always been. People.
This post was edited on 5/27/25 at 10:30 am
Posted on 5/27/25 at 10:33 am to theunknownknight
quote:
The danger is what it has always been. People.
Yes, root cause of the danger. We’ll see.
Posted on 5/27/25 at 10:34 am to RaoulDuke504
quote:
So you don’t consider that a threat the way AI is advancing, just a few months ago AI had extra fingers and couldn’t talk. You think it won’t advance in that area as well?
I guess it depends on what you mean by threat and who its threatening. I dont know if I would consider it to be the direct threat. Its just trying to sort out its programming and we are seeing the results of some kind of misalignment by the programmers.. Its not like its gaining sentience if that is what your are fearing. That doenst mean that misalignment cant have disastrous consequences even when dealing with less than true AGI systems, but learning these flaws is part of the process.
Posted on 5/27/25 at 11:04 am to wareaglepete
quote:
Here come the code nerds with their, it’s the programming.
What else what it be?
Posted on 5/27/25 at 11:06 am to RaoulDuke504
quote:that's all great, but the real issue I have with them is that Chat GPT won't tell me any dirty jokes because it's against its policies...
OpenAI refused to shut it self down when commanded, also attempted to create backdoors.
Page 1 of 1
Popular
Back to top
Follow TigerDroppings for LSU Football News