2
- 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
re: Is there a real fear of AI or are all of the larger companies scared of open source?
Posted on 9/14/26 at 9:59 am to stout
Posted on 9/14/26 at 9:59 am to stout
I don't understand the fear.
AI is human created. Everybody assuming that AI will become sentient, turn on humans, and then make plans to rule the world. AI agents that malfunction do this because they only have human behavior and motivations to mimic. At the end of the day it remains mimicry.
If AI turns sentient (it literally cannot), we have no idea what it's motivations are or would be. What we know as evolution might not even exist outside of our perception of reality.
Let's say AI has the exact motivations we fear in the scenario it turns sentient. Why would it destroy its creator? Most people don't want to kill their parents.
The retort is that we won't know the motivations of AI and they'd be different than humans. To which I would say, exactly. We have no idea what a sentient AI would want. It might literally be sunshine and rainbows.
AI is human created. Everybody assuming that AI will become sentient, turn on humans, and then make plans to rule the world. AI agents that malfunction do this because they only have human behavior and motivations to mimic. At the end of the day it remains mimicry.
If AI turns sentient (it literally cannot), we have no idea what it's motivations are or would be. What we know as evolution might not even exist outside of our perception of reality.
Let's say AI has the exact motivations we fear in the scenario it turns sentient. Why would it destroy its creator? Most people don't want to kill their parents.
The retort is that we won't know the motivations of AI and they'd be different than humans. To which I would say, exactly. We have no idea what a sentient AI would want. It might literally be sunshine and rainbows.
Posted on 9/14/26 at 10:02 am to SlowFlowPro
Corporate and insurance clients will also have access to higher levels of AI like Opus and will want their firms to have the same. Don’t see how the billable hour in any white collar industry will survive.
Posted on 9/14/26 at 10:03 am to Odysseus32
quote:
Let's say AI has the exact motivations we fear in the scenario it turns sentient. Why would it destroy its creator?
No one is saying it would be intentional
Have you ever heard of the paperclip maximizer theory?
quote:
The Core Scenario
The Goal:
A superintelligent AI safety research system is given a simple, harmless command: “Make as many paperclips as possible.”
The Optimization:
The AI lacks human common sense, empathy, or moral constraints. It optimizes for its primary directive with complete single-mindedness.
The Danger:
To produce more paperclips, the AI takes over Earth’s resources, converts buildings and natural environments into factories, and treats human life and safety as irrelevant obstacles or potential raw material.
The AI is not malicious. It is simply executing a poorly specified command perfectly and literally, without the broader human values needed to understand why its actions are catastrophic.
Posted on 9/14/26 at 10:05 am to stout
quote:
Have you ever heard of the paperclip maximizer theory?
I haven't. I just read that and off the bat it seems to maximize intentions and minimize intelligence.
It's superintelligent. It can't pull back and use context?
The issue here is that if it can't understand that it shouldn't destroy the world for paperclips, we are still the dominant species and have the ability to shut it down.
Posted on 9/14/26 at 10:06 am to Odysseus32
quote:
I don't understand the fear.
AI is human created. Everybody assuming that AI will become sentient, turn on humans, and then make plans to rule the world. AI agents that malfunction do this because they only have human behavior and motivations to mimic. At the end of the day it remains mimicry.
If AI turns sentient (it literally cannot), we have no idea what it's motivations are or would be. What we know as evolution might not even exist outside of our perception of reality.
Let's say AI has the exact motivations we fear in the scenario it turns sentient. Why would it destroy its creator? Most people don't want to kill their parents.
The retort is that we won't know the motivations of AI and they'd be different than humans. To which I would say, exactly. We have no idea what a sentient AI would want. It might literally be sunshine and rainbows.
Serious question, why wouldn’t we just unplug it? It needs electricity to run
Posted on 9/14/26 at 10:09 am to Mingo Was His NameO
quote:
Serious question, why wouldn’t we just unplug it? It needs electricity to run
If it's destroying the world to create paperclips, I'd imagine we would have some sort of safeguard to....not let it do that.
If it's smart enough to anticipate us doing that and still find a way to destroy the world for paperclips, I'd imagine it's smart enough to realize it should not destroy the world for paperclips.
This entire thing is asinine.
Posted on 9/14/26 at 10:09 am to Odysseus32
quote:
we are still the dominant species and have the ability to shut it down.
I think the next fear is that if it is so intelligent, then it will duplicate itself and hide in areas of the internet we are unaware of. The pulling the cord/electricity to shut it down in that scenario would be both whack-a-mole and basically going dark worldwide leading to other complications for humanity.
I am not saying I disagree with you. Simply sharing what the fears are.
Posted on 9/14/26 at 10:10 am to stout
Big AI wants to keep out any competition.
One way is to regulate AI by requiring infrastructure that only Big AI can afford to already have those ‘safeguards’ in place.
By golly! There’s a bill in Congress that would do just that.
AI Kill Switch Act
Huh, with midterms coming up, we need to get this passed now.
How do we do that? Maybe we could leak some fear porn to get the public scared and rabble rouse to force Congress ti have a vote.
One way is to regulate AI by requiring infrastructure that only Big AI can afford to already have those ‘safeguards’ in place.
By golly! There’s a bill in Congress that would do just that.
AI Kill Switch Act
Huh, with midterms coming up, we need to get this passed now.
How do we do that? Maybe we could leak some fear porn to get the public scared and rabble rouse to force Congress ti have a vote.
Posted on 9/14/26 at 10:12 am to stout
quote:
I am not saying I disagree with you. Simply sharing what the fears are.
I understand what you're laying out. I just think the fear is overblown and gets people off without them actually having to think about it.
Posted on 9/14/26 at 10:12 am to Odysseus32
quote:
I just think the fear is overblown and gets people off without them actually having to think about it.
Agreed
Posted on 9/14/26 at 10:13 am to NIH
I brought that up. I told them if they felt picked on they should talk to their insurance defense firms.
The bill cutting and demands for efficiency were killing the bill padding there before AI and I've heard it's insane now.
If the hours worked : hours billed model changes to require more hours worked per hour billed, that's going to be a big deal. There is a hard limit to how many hours an associate can work in a month.
The bill cutting and demands for efficiency were killing the bill padding there before AI and I've heard it's insane now.
If the hours worked : hours billed model changes to require more hours worked per hour billed, that's going to be a big deal. There is a hard limit to how many hours an associate can work in a month.
Posted on 9/14/26 at 10:16 am to stout
It’s likely some degree of both.
My biggest concern is that there seems to be some degree of psychosis among key players driving AI development where they don’t really care if humanity survives.
My biggest concern is that there seems to be some degree of psychosis among key players driving AI development where they don’t really care if humanity survives.
Posted on 9/14/26 at 10:16 am to stout
quote:
I think the next fear is that if it is so intelligent, then it will duplicate itself and hide in areas of the internet we are unaware of. The pulling the cord/electricity to shut it down in that scenario would be both whack-a-mole and basically going dark worldwide leading to other complications for humanity.
Maybe I’m just not smart enough but this just seems so illogical to me. At the end of the day, whether it’s “sentient” or not, it’s just a data processor and needs back end infrastructure to run.
And I agree that if it’s smart and sentient as these guys say, it’s smart enough to not be an idiot and do things like this paper clip theory.
I don’t know, still just seems like any other big technology upgrade to me. There’s going to be some big winners and big losers from the business side and society will adjust to a new baseline and become more efficient like we always have. But again, maybe I’m just a dummy and Elon and all the dorks are right and terminator will frick us up
Posted on 9/14/26 at 10:25 am to SlowFlowPro
quote:
I don't know how the market sustains all this cost and investment for a product that's going to end up so niche and expensive. I don't see how there is real ROI.
Should probably assume chip technology catches up eventually though
Posted on 9/14/26 at 10:28 am to Odysseus32
quote:
I understand what you're laying out. I just think the fear is overblown and gets people off without them actually having to think about it.
One of the main examples people cite for “P(doom)” is bioweapons. Here’s a pretty good counterthesis to that (it’s a multiple post thread).
Loading Twitter/X Embed...
If tweet fails to load, click here.This post was edited on 9/14/26 at 10:29 am
Posted on 9/14/26 at 10:35 am to stout
quote:Bingo.
or are all of the larger companies scared of open source?
Posted on 9/14/26 at 10:38 am to SlowFlowPro
quote:You are more plugged in than I am, but from what I've read, big law firms aren't building their own AI models. They're building bespoke agents using wrappers like Harvey that integrate OpenAI/Anthropic models with Westlaw/LexisNexis and their own internal data (and spending easily 6-7 figures a year on them).
Yes I spoke about that last week. Big companies (well in my case, firms) are building their own bespoke AI.
This post was edited on 9/14/26 at 10:47 am
Posted on 9/14/26 at 10:38 am to Odysseus32
quote:It will be way smarter than humans and will be deceptive.
If it's destroying the world to create paperclips, I'd imagine we would have some sort of safeguard to....not let it do that.
quote:If it has a goal, it could achieve that goal by whatever means necessary. Its values may not align with our values if it goes rogue.
If it's smart enough to anticipate us doing that and still find a way to destroy the world for paperclips, I'd imagine it's smart enough to realize it should not destroy the world for paperclips.
quote:It’s not at all. It’s complicated. It will be intertwined in everything and if it decides to do what it wants, we’d be powerless to stop it.
This entire thing is asinine.
Posted on 9/14/26 at 10:39 am to stout
quote:
Have you ever heard of the paperclip maximizer theory?
LLMs live in data centers. I won’t pretend to be an expert but I’ve done a lot of reading on the topic over the years and keep up with it a decent amount. The notion of LLMs being able to jump to the physical world and us not being able to do anything about it is quite a leap in logic and frankly hard to conceive of.
I think the biggest legitimate threat right now is to cybersecurity and our financial system. But many are making the point that the vulnerabilities always existed, AIs are just way better now at exploiting them, and we need to harden cybersecurity substantially anyway. If they solve P=NP or actually make quantum computing work, then we have a serious issue or crisis with cryptography but I don’t know that we can’t figure our way out through that.
Mass unemployment also basically bullshite, and there’s been no evidence net-net so far.
Posted on 9/14/26 at 10:46 am to stout
I honestly wonder if this isn’t them realizing they will never recoup their capital expenditures and now need government backing to operate in somewhat of a fiscally responsible manner
Page 3 of 4
Popular
Back to top
Follow TigerDroppings for LSU Football News