Topic: The tech bros think being eccentric assholes is a sign of genius.
Anonymous A started this discussion 3 weeks ago #135,437
It mistakes correlation with causation and lets their douche flag fly. Zuckerberg with his Jew fro and shrooms dealer outfit and BJJ private lessons. Elon being an uncouth bigot who never says anything notable. Larry Ellison with his Hawaiian island and ERP software wealth that qualifies him as a social Darwinist.
It’s now the nation’s problem that nobody wanted to fuck these guys in high school so they lost themselves in Atlas Shrugged and Middle Earth.
It’s now the nation’s problem that nobody wanted to fuck these guys in high school so they lost themselves in Atlas Shrugged and Middle Earth.
Mr. Big Black Friday joined in and replied with this 3 weeks ago, 52 minutes later[^] [v] #1,439,194
raping nigguhs!
Anonymous C joined in and replied with this 3 weeks ago, 2 hours later, 3 hours after the original post[^] [v] #1,439,239
Anonymous D joined in and replied with this 3 weeks ago, 23 hours later, 1 day after the original post[^] [v] #1,439,451
AI investment is a scam
https://www.wheresyoured.at/ai-is-slowing-down/
https://www.wheresyoured.at/ai-is-slowing-down/
Anonymous E joined in and replied with this 3 weeks ago, 32 minutes later, 1 day after the original post[^] [v] #1,439,472
@previous (D)
You don't have any money to invest anyway, stop acting like you are doing something important by reading slop articles.
You don't have any money to invest anyway, stop acting like you are doing something important by reading slop articles.
Anonymous F joined in and replied with this 3 weeks ago, 2 minutes later, 1 day after the original post[^] [v] #1,439,474
@previous (E)
It basically is a scam though. OpenAI isn’t profitable, they just burn through money selling a product for less money than it costs to produce while running a Ponzi scheme where investors give them money for shares in the hopes another investor will come along and pay them more for their shares, when the company isn’t even making money in the first place. The entire American economy is propped up by trillion dollar scams.
It basically is a scam though. OpenAI isn’t profitable, they just burn through money selling a product for less money than it costs to produce while running a Ponzi scheme where investors give them money for shares in the hopes another investor will come along and pay them more for their shares, when the company isn’t even making money in the first place. The entire American economy is propped up by trillion dollar scams.
Anonymous G joined in and replied with this 3 weeks ago, 8 minutes later, 1 day after the original post[^] [v] #1,439,477
@previous (F)
Then invest in something else. But land or gold or a stake in a business you understand.
Stop worrying about how other people invest.
Then invest in something else. But land or gold or a stake in a business you understand.
Stop worrying about how other people invest.
Anonymous F replied with this 3 weeks ago, 1 minute later, 1 day after the original post[^] [v] #1,439,480
Anonymous G replied with this 3 weeks ago, 44 seconds later, 1 day after the original post[^] [v] #1,439,481
Anonymous F replied with this 3 weeks ago, 37 seconds later, 1 day after the original post[^] [v] #1,439,483
Anonymous G replied with this 3 weeks ago, 41 seconds later, 1 day after the original post[^] [v] #1,439,484
@previous (F)
You were the one that wrote out a block of text telling people not to invest, I said it doesn't matter.
Keep up!
You were the one that wrote out a block of text telling people not to invest, I said it doesn't matter.
Keep up!
Anonymous F replied with this 3 weeks ago, 47 seconds later, 1 day after the original post[^] [v] #1,439,485
Anonymous G replied with this 3 weeks ago, 4 minutes later, 1 day after the original post[^] [v] #1,439,486
@previous (F)
> When did I tell people not to invest?
You called it a scam:
@1,439,474 (F)
> It basically is a scam though.
> When did I tell people not to invest?
You called it a scam:
@1,439,474 (F)
> It basically is a scam though.
Anonymous F replied with this 3 weeks ago, 35 seconds later, 1 day after the original post[^] [v] #1,439,487
Anonymous H joined in and replied with this 3 weeks ago, 52 seconds later, 1 day after the original post[^] [v] #1,439,488
I’m curious how it shakes out. It’s going to be a big bust and regrouping. The 2000 dot com flameout on steroids, and obviously the Internet didn’t perish.
There’s a ton of pressure from Wild West open source models and some of the leaders are following their dodgy security protocols to stay ahead.
Multiple customers of mine have had issues with Mythos actively rewriting security protocols to please the user.
There’s a ton of pressure from Wild West open source models and some of the leaders are following their dodgy security protocols to stay ahead.
Multiple customers of mine have had issues with Mythos actively rewriting security protocols to please the user.
Anonymous F replied with this 3 weeks ago, 17 seconds later, 1 day after the original post[^] [v] #1,439,489
(I’ll give you a hint: it’s a trick question because I never told people not to invest you knob.)
Anonymous F double-posted this 3 weeks ago, 1 minute later, 1 day after the original post[^] [v] #1,439,490
@1,439,488 (H)
Ngl, I was programming with some open source LLMs recently, it’s incredible how when you modify the history you can convince these things absolutely anything is true. It annoys me when people say it’s intelligent when it’s so clearly not.
Ngl, I was programming with some open source LLMs recently, it’s incredible how when you modify the history you can convince these things absolutely anything is true. It annoys me when people say it’s intelligent when it’s so clearly not.
Anonymous G replied with this 3 weeks ago, 2 minutes later, 1 day after the original post[^] [v] #1,439,492
Anonymous F replied with this 3 weeks ago, 40 seconds later, 1 day after the original post[^] [v] #1,439,494
Rather than try and explain it I’ll just leave a video of somebody else who basically did the same thing so you can see what I’m talking about.
https://youtu.be/WP5_XJY_P0Q
https://youtu.be/WP5_XJY_P0Q
Anonymous G replied with this 3 weeks ago, 28 minutes later, 1 day after the original post[^] [v] #1,439,497
@previous (F)
The API treats every call independently, therefore LLMs aren't "intelligent", which you don't precisely define.
What consequence do you think there is to any of this?
I don't see any errors in his video really, but what does any of it change? Nothing.
It's not a complete thought. "Oh this is different, which means it doesn't count, and that matters because..."
The API treats every call independently, therefore LLMs aren't "intelligent", which you don't precisely define.
What consequence do you think there is to any of this?
I don't see any errors in his video really, but what does any of it change? Nothing.
It's not a complete thought. "Oh this is different, which means it doesn't count, and that matters because..."
Anonymous E replied with this 3 weeks ago, 2 hours later, 1 day after the original post[^] [v] #1,439,510
Mr. Bloody Lemonade joined in and replied with this 3 weeks ago, 35 minutes later, 1 day after the original post[^] [v] #1,439,517
@1,439,497 (G)
> The API treats every call independently, therefore LLMs aren't "intelligent", which you don't precisely define.
>
> What consequence do you think there is to any of this?
>
> I don't see any errors in his video really, but what does any of it change? Nothing.
>
> It's not a complete thought. "Oh this is different, which means it doesn't count, and that matters because..."
This is such a weird reaction.
> The API treats every call independently, therefore LLMs aren't "intelligent", which you don't precisely define.
>
> What consequence do you think there is to any of this?
>
> I don't see any errors in his video really, but what does any of it change? Nothing.
>
> It's not a complete thought. "Oh this is different, which means it doesn't count, and that matters because..."
This is such a weird reaction.
Mr. Bloody Lemonade double-posted this 3 weeks ago, 3 minutes later, 1 day after the original post[^] [v] #1,439,518
An LLM is just a stateless function. You have an input, you have an output, the input determines the output. You can add some randomness in, it’s a completely deterministic stateless system. There’s no intelligence there, the only way you can think an LLM is intelligent is if you either don’t understand the technology at all or are completely delusional.
Anonymous J joined in and replied with this 3 weeks ago, 3 minutes later, 1 day after the original post[^] [v] #1,439,519
Anonymous G replied with this 3 weeks ago, 20 seconds later, 1 day after the original post[^] [v] #1,439,520
Anonymous G double-posted this 3 weeks ago, 1 minute later, 1 day after the original post[^] [v] #1,439,521
@1,439,518 (Mr. Bloody Lemonade)
It's not deterministic, the same input will give different outputs by chance for the exact same model.
And before you respond to this ask yourself if your new idea would also apply to a physical human brain, thanks.
It's not deterministic, the same input will give different outputs by chance for the exact same model.
And before you respond to this ask yourself if your new idea would also apply to a physical human brain, thanks.
Mr. Bloody Lemonade replied with this 3 weeks ago, 4 minutes later, 1 day after the original post[^] [v] #1,439,528
@previous (G)
You do realize that computers can’t generate random numbers, right? There are only two ways computers can generate "random" numbers: PRNGs (which are statistically random but not actually random), and you can also generate random values by using sensor to get randomness from the environment. But actual programs can’t generate actually random outputs without random inputs. The only reason why you get a different result out of the LLM is because your computer is either using a PRNG underneath and is generating statistically random yet deterministic values or it’s using input devices to generate random numbers in some way from environmental noise. But if you gave an LLM the same chat history and prompt as an input with the same PRNG algorithm with the same seed it would generate exactly the same output.
You do realize that computers can’t generate random numbers, right? There are only two ways computers can generate "random" numbers: PRNGs (which are statistically random but not actually random), and you can also generate random values by using sensor to get randomness from the environment. But actual programs can’t generate actually random outputs without random inputs. The only reason why you get a different result out of the LLM is because your computer is either using a PRNG underneath and is generating statistically random yet deterministic values or it’s using input devices to generate random numbers in some way from environmental noise. But if you gave an LLM the same chat history and prompt as an input with the same PRNG algorithm with the same seed it would generate exactly the same output.
Anonymous G replied with this 3 weeks ago, 1 minute later, 1 day after the original post[^] [v] #1,439,529
@previous (Mr. Bloody Lemonade)
Given the challenge, do you have some proof human neurons do generate random numbers?
Given the challenge, do you have some proof human neurons do generate random numbers?
Anonymous J replied with this 3 weeks ago, 2 minutes later, 1 day after the original post[^] [v] #1,439,531
@1,439,528 (Mr. Bloody Lemonade)
> You do realize that computers can’t generate random numbers, right?
skill issue
> You do realize that computers can’t generate random numbers, right?
skill issue
Mr. Bloody Lemonade replied with this 3 weeks ago, 24 seconds later, 1 day after the original post[^] [v] #1,439,532
The difference between actually random numbers is actually random numbers aren’t deterministic, which means you can’t predict them. Statistically random numbers are just sequences of numbers that generally produce values that are spread evenly as if they were actually random, so you can use statistically random numbers in algorithms that rely on randomness that don’t actually have to be cryptographically secure. For example, the pagerank algorithm websites use utilizes a damping factor that determines how often it will jump to a random link.
Mr. Bloody Lemonade double-posted this 3 weeks ago, 1 minute later, 1 day after the original post[^] [v] #1,439,534
@1,439,529 (G)
If human neurons utilize quantum mechanics in some way, it might be possible that they do. How neurons work isn’t fully understood and perceptrons in artificial neural networks are based on an (inaccurate) understanding of how biological neurons work, but it is still a very powerful model, even if it is technically wrong.
If human neurons utilize quantum mechanics in some way, it might be possible that they do. How neurons work isn’t fully understood and perceptrons in artificial neural networks are based on an (inaccurate) understanding of how biological neurons work, but it is still a very powerful model, even if it is technically wrong.
Mr. Bloody Lemonade triple-posted this 3 weeks ago, 3 minutes later, 1 day after the original post[^] [v] #1,439,535
It is generally accepted though that an individual biological neuron can do tasks that take multiple perceptrons so one perceptron isn’t equivalent to one neuron.
Anonymous G replied with this 3 weeks ago, 15 minutes later, 1 day after the original post[^] [v] #1,439,541
@1,439,534 (Mr. Bloody Lemonade)
So no, you can't show that human neurons generate random numbers but its a problem that silicon doesn't either.
So no, you can't show that human neurons generate random numbers but its a problem that silicon doesn't either.
Anonymous G double-posted this 3 weeks ago, 2 minutes later, 1 day after the original post[^] [v] #1,439,542
@1,439,535 (Mr. Bloody Lemonade)
> one perceptron isn’t equivalent to one neuron.
Ok, great, who said it was?
The fact is you have two systems thatprocess information, and you decided one counts as intelligence and one doesn't. You have neither some evidence neurons are doing something special nor any explanation for why your determination that one is really intelligence would matter.
Which means you've done nothing at all really. You took a meaningless label without any litmus test or consequences and applied it to one and not another arbitrarily.
> one perceptron isn’t equivalent to one neuron.
Ok, great, who said it was?
The fact is you have two systems thatprocess information, and you decided one counts as intelligence and one doesn't. You have neither some evidence neurons are doing something special nor any explanation for why your determination that one is really intelligence would matter.
Which means you've done nothing at all really. You took a meaningless label without any litmus test or consequences and applied it to one and not another arbitrarily.