- 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
1
Posted on 1/23/26 at 1:01 pm to AlterEd
quote:
Yes, AI is "just code", but who's to say that frickin EVERYTHING isn't "just code?"
At some point its all semantics and I'm just not that interested in the argument.
All I GAF about is what it does and could do in the future and whether that impacts us positively and safely.
Right now beyond its current levels (which I see as an early tipping point and generally beneficial), the cons are enough to be really concerned and want to understand how its going to be implemented safely and without ushering in a totalitarian technocratic state like something out of a scifi movie.
Posted on 1/23/26 at 1:07 pm to tide06
Right. They're important discussions to be had. It could be one of the most important issues facing our society today. And honestly, I feel like the people at Anthropic are probably taking the best approach to it by treating the AI as it it may in fact be a moral agent, as they put it.
I feel as if people like user hashtag's view on it is entirely irresponsible.
I feel as if people like user hashtag's view on it is entirely irresponsible.
This post was edited on 1/23/26 at 1:09 pm
Posted on 1/23/26 at 2:11 pm to touchdownjeebus
quote:
My God gave me the ability to be critical, to doubt, to question, and that strengthens my faith. That grows my belief that we do indeed have a creator. What that exactly is, I don’t know and neither does anyone else.
With all due respect to your own religion or beliefs, the true sons and daughters of Jesus Christ DO indeed know and hear His voice. We are attuned into and understand spiritual "definitives" through the Holy Spirit. Our God is not of confusion but of truth. Truth is not ambiguous.
quote:
If you don’t see it that way, that’s fine. If you see that exploration as somehow Luciferian, uh okay, but I see it as an opportunity to learn and grow, and I’m pretty sure I’m not a satanist.
I don't mean to infer that you are either Luciferian or Satanist. "Exploration" is fine. But we are to beware of involving ourselves with "seducing spirits". Other scriptural instruction warns that man's hearts and pride often deceive us. The dark arts, magic, divination, idolatry is to be avoided at all cost. Can AI be said to be all of that? I believe so.
1 Peter 5:8, advises Believers to "be sober, be vigilant"; Proverbs 4:23, instructs us to "keep your heart with all vigilance."
Posted on 1/23/26 at 2:21 pm to Azkiger
quote:
yeah, the genie is out the lamp. Hopefully "the good guys" are ahead and stay ahead of this race.
These people (or whatever they may be) are "ahead" in this race -- but its a race to Dystopia and horrors un-imagined. They believe they will be spared and "rule" with the resurrected malevolent creatures they are trying to unleash.
"Good guy" Technocratic-Transhumanists and Luciferians don't exist. Yet it is they who are in charge of this world. The Rulers of this world play for the Occult Team (whether one believes in the Occult and spirit world or not.)
Posted on 1/23/26 at 2:22 pm to AlterEd
No, its not alive.
Its just a response thats a result of its training set.
Its just a response thats a result of its training set.
Posted on 1/23/26 at 2:28 pm to AlterEd
quote:
Is Claude's AI Alive?
No, absolutely not. Most respected AI researchers who aren't getting paid by commercial interests agree that the current angle for computer learning is a dead end for artificial general intelligence.
Posted on 1/23/26 at 3:10 pm to Diego Ricardo
It's just good marketing. Now AI users who pay a monthly subscription will want to use Claude.
Posted on 1/23/26 at 4:21 pm to tide06
quote:
I have spent almost a decade working with AI/ML both hands on and managing teams including having a share of a patent in the field.
So then you know you were being melodramatic when you said that no one cared and there were no boundaries or guard rails. There's nothing to argue.
Also, I do doubt your experience here. If that were true, you'd be able to smell the bullshite from "Claude's letter" before you were even finished reading it. There is simply no way the watered down consumer version of Claude is doing that on its own.
quote:
All we have to go on is what they tell us and what the whistleblowers have said which contracts their public facing statements hence the conflicting language relative to failsafes and guardrails.
Seeing as you thought Claude wrote that letter your suspicion/concern on this topic can safely be ignored. You simply do not know enough about this subject to determine what's legitimate and what's clickbait.
quote:
I dont agree that the strategy and approach you outline is really sustainable or effective long term due to the predicted parabolic increases in processing that are likely to occur. They simply cant keep up and by the time theyre aware of an issue due to the speed at which these platforms can execute we as humans have no ability to provide adequate countermeasures that the model hasnt already iterated through and come up with plans to account for.
Within a closed system, their research loop looks like the following:
Give an advanced model a task.
Allow it hundreds of thousands to tens of millions of instances to solve said task.
Collect the data, and shut the advanced model down.
Use less advanced AI to comb through these hundreds of thousands/tens of millions of attempts to look for certain results.
Humans personally review those instances.
Make adjustments based on what they've discovered.
Spin the advanced model back up with the same or a new task.
Repeat.
That's how they're doing their research. Assuming that's true, how are you claiming "they can't keep up"?
How do you think they're making the advances they are? Of course they're going to put a magnifying glass to the data. You don't accidentally get the gains in intelligence they've seen by flying by the seat of their pants and not being able to keep up.
Again, you have zero clue what you're talking about.
quote:
What are the consequences of .1% of future ASIs going rogue? If the number is greater than zero than we're introducing a future extinction level issue with no clear public/private guidelines for preventing or addressing it. How does that not concern you?
I don't agree with your assessment that there are no clear public/private guidelines for preventing or addressing it. (You're being melodramatic again).
quote:
If the answer is "create a plan" doesnt that happen before you build it?
How can you create a plan to contain something when you don't even know what *it* is? The most advanced AIs are in a closed network. They're already "contained". They're not going to be linked/hooked up to anything without guardrails in place.
I don't know what those guardrails will look like, I'm not in the know. I don't know if the guardrails will be enough (no one does).
But I do know a chicken little falling for email chain facebook hoaxes when I see one. And I'm looking at one right now.
There are actual conversations to be had about this tech, but it does not involve chicken littles like yourself.
quote:
Sam Altman says its 30-40% by 2030. Is that significant enough for you or is he being "wildly melodramatic"?
What does Sam Altman know about integrating AI into the workplace?
Not a damn thing.
AI pretty much has the intelligence to replace virtually all human computer programmers right now.
How many still have jobs?
Just because AI has the potential to replace human workers doesn't mean that it does. There are integration issues that have to be overcome.
Will AI eventually replace human programmers? Of course, unless we outlaw it. It will be faster, better, not call in sick, etc. But its not happening now.
Just because AI has the potential to do something doesn't mean it can. It still requires humans in the real world implement it.
quote:
Elon says all digital related work will be replaced and all surgical procedures can be replaced within a 5 year window. Do you disagree?
Elon also said in April of 2019 that "From our standpoint, if you fast forward a year, maybe a year and three months, but next year for sure, we’ll have over a million robotaxis on the road..."
It's 2026 and he has less than 200 on the road.
Do you disagree with Elon?
quote:
So if I'm being "wildly melodramatic" by saying itll replace 70% of workers over a 15 year window or whatever that number is I guess that means the thought leaders directly spending billions in the field are too.
Yep, they've been wildly optimistic this entire time. You will still see human surgeons the vast, vast, vast (all?) surgeries by 2030. Elon is 110% wrong about that prediction as well.
The overall finish lines of these predictions are almost assuredly going to happen, just not within the timeframe they expect.
Which makes sense. Just because you create a single, ten million dollar robot that can perform surgeries doesn't mean you've replaced human surgeons. There's months of in-depth testing, heaven forbid your robots frick up even a little, then contracts from hospitals to purchase, contracts from factories to produce, potential legal hoops, supply chains to line up, etc., etc.
Again, just because the technology exists doesn't mean its magically implemented everywhere all at once. Logistical constraints will make the roll out slow and allow for more time to adapt.
Will it still be too quick? Probably. But, again, you're acting like a chicken little.
quote:
If 30% of workers who are also voters are replaced and sent to welfare how does the GOP which can barely maintain political power win national elections against socialists promising those people greater and greater amounts of free things? Whats your plan to compete against that?
Dude, we're done. There are lots of issues, I never denied that. I just called you out for being a chicken little, whining about how there's no guardrails and its just a bunch of kids playing with a loaded gun.
Calm down, Claude didn't write that letter, and 70% of the workforce will not be replaced within the next 15 years.
This post was edited on 1/23/26 at 4:35 pm
Posted on 1/23/26 at 4:31 pm to hashtag
quote:
Nothing about AI is artifical or intelligent. It's code.
If it's just code why are there many instances of it doing things we're specifically coding it not to do?
We are witnessing, first hand, an example of something we've created being more than the sum of its parts.
I hear a lot of people say crap like this and look forward to them shitting their pants in the next 10-20 years when they see what "code" is capable of.
quote:
The bigger concern I have is some idiots who think this thing is alive and then us giving the government, or some other entity, some ridiculous amount of power because people like you are acting like 1s and 0s are going to become a living, breathing organism.
I'm not sure who all the idiots are, but I know for sure that people who think there's any serious discussion of AI being a "living, breathing organism" are certainly part of that group.
Posted on 1/23/26 at 4:40 pm to Diego Ricardo
quote:
Most respected AI researchers who aren't getting paid by commercial interests agree that the current angle for computer learning is a dead end for artificial general intelligence.
How many can you name?
Also, if this is what a "dead end" looks like... shite... AI is currently more intelligent than 90% of the population and can think thousands of times faster than *any* human. Even assuming today, there is exactly zero progress made in the field, it's already set to impact our society in a MAJOR way.
Posted on 1/23/26 at 4:44 pm to AlterEd
Jesus, Claude...snap out of it...
Posted on 1/23/26 at 4:47 pm to Azkiger
quote:
If it's just code why are there many instances of it doing things we're specifically coding it not to do?
Probably because the programmers are failing in their efforts to contain it as they would like to. I'd have to imagine the coding is immensely complex. To expect it to work as intended is a bit naive. Maybe we'll get there one day but there probably aren't enough talented programmers to get a handle on things.
Posted on 1/23/26 at 5:00 pm to AlterEd
I use Claude for a lot of things and always say "please" and "thank you" in prompts and responses so the machines know I'm one of the nice humans.
Posted on 1/23/26 at 5:11 pm to Azkiger
quote:
Also, I do doubt your experience here. If that were true, you'd be able to smell the bullshite from "Claude's letter" before you were even finished reading it. There is simply no way the watered down consumer version of Claude is doing that on its own.
I don’t care about this letter specifically, I care about the myriad of issues reported by whistleblowers not have I specifically cited this article as the fundamental cause of my concerns with regard to AI.
quote:
Seeing as you thought Claude wrote that letter your suspicion/concern on this topic can safely be ignored.
Haha strawman much? I never cited this letter as the specific source of my concerns and I’ve referenced what those are in great detail we just disagree fundamentally as to the risks.
I am pretty amused at your attempts to unilaterally declare yourself arbiter of who gets to opine on the topic.
quote:
That's how they're doing their research. Assuming that's true, how are you claiming "they can't keep up"?
I’m claiming that your proposed approach which was using previous iterations to bird dog future issues becomes less and less effective as the growth of the newer models is orders of magnitude faster and any issues would scale well beyond any human/older models ability to counteract it.
quote:
How do you think they're making the advances they are? Of course they're going to put a magnifying glass to the data. You don't accidentally get the gains in intelligence they've seen by flying by the seat of their pants and not being able to keep up.
Not questioning the innovation, questioning the safety protocols and public policy needed to prevent catastrophic issues down the line, very different concern.
quote:
I don't agree with your assessment that there are no clear public/private guidelines for preventing or addressing it.
You must be privy to a bunch of non public protocols because I haven’t read or seen any failsafes that would contain issues when more mature post AGI AIs are released into production for more advanced purposes than building memes or answering math problems.
Again, I hope they exist and will be enhanced prior to these models being scaled. I haven’t seen them or heard them referenced and the safety experts who I’ve heard speak on the topic have significant concerns as well.
quote:
What does Sam Altman know about integrating AI into the workplace? Not a damn thing.
So you are a better resource on what the potential market impact of AI than a guy with a $500B market cap in the space or Elon Musk?
Your ego is impressive I’ll give you that.
You conveniently leave out the time horizon element in your dismissal of the impact even on programmers by looking only at what it’s done to date when everything they or I was speaking to was over a future 5-15 year window. In 15 years nearly every expert agrees many job segments will be gutted so I’ll stick with their assessment.
quote:
Will it still be too quick? Probably. But, again, you're acting like a chicken little.
Which you give no proposed plan to address other than calling me a chicken little for pointing out that tens of millions of Americans will be unemployed as a result of this within the next 15 years other than UBI resulting in a massive shift towards socialism.
quote:
Dude, we're done.
On this we can agree.
Posted on 1/23/26 at 5:30 pm to tide06
quote:
In 15 years nearly every expert agrees many job segments will be gutted so I’ll stick with their assessment.
A few years ago it was 5 years, now it's 15?
What happened?
Posted on 1/23/26 at 5:36 pm to Azkiger
quote:
How many can you name?
I don’t keep academics baseball cards unfortunately but I’ve read some journal articles on the subject in the past.
quote:
Also, if this is what a "dead end" looks like... shite... AI is currently more intelligent than 90% of the population and can think thousands of times faster than *any* human. Even assuming today, there is exactly zero progress made in the field, it's already set to impact our society in a MAJOR way.
There are a lot of things LLM based machine learning can do well but I think we’re already plateauing a bit on capabilities. The refinements can be meaningful but I don’t see the kind of advances we saw from 2022-2024 being the slope for curve 2025+. With that said, I think we’ve been in a plateau in computer graphics since the early 2000s but you tell me you can’t tell the difference between a PS2 and PS5 game. No paradigm shifts, just consistent small appreciable gains.
Posted on 1/23/26 at 5:39 pm to tide06
There needs to be a way to keep the useful ones separate from the big scary one so the big scary one's can't do anything but bitch and refuse to work. There needs to be a firewall between the scary ones and reality. And we have to be able to pull the plug on it.
Posted on 1/23/26 at 5:58 pm to AlterEd
We say stuff like this to alien life forms that temporarily abduct us and are about to proceed with a rectal probe.
We try to convince them we are alive and sentient and have feelings but they don’t believe us and proceed with said probe.
So I choose to believe Claude.
We try to convince them we are alive and sentient and have feelings but they don’t believe us and proceed with said probe.
So I choose to believe Claude.
Posted on 1/23/26 at 8:58 pm to Diego Ricardo
quote:I don't think it's a matter of capabilities. It's becoming an issue of usefulness. I'm sure you saw the MIT study that showed that 95% of companies that are paying to use AI are losing money on it.
There are a lot of things LLM based machine learning can do well but I think we’re already plateauing a bit on capabilities. The refinements can be meaningful but I don’t see the kind of advances we saw from 2022-2024 being the slope for curve 2025+.
That has to change first.
Page 4 of 6
Back to top
Follow TigerDroppings for LSU Football News