Machine Learning and Artificial Intelligence Thread

[PurpleUrkel]

If we're honest, we were always separated from nature, it's just an arbitrary definition once we attain greater control over nature.

We have no control over nature. Zero. I'm talking about the spiritual rejuvination that comes from the peace and quiet of being in nature. God's creation of photosynthesis, when observed "properly," which means not looking at it scientifically of for "answers," can bring a great sense of mysterious wonder and peace to the human soul. Sure, one can do this starring at one's house plants in a Manhatten penthouse but the effect is much stronger and more profound when in a huge forested ecosystem untouched by man.

Humans used to barely live until 30...

That is an illusion of the scientific (which is often demonic) mind. But seeing as how we are talking perceivable science here, and for those who have such a deep need for control that they frame everything in scientific and measurable terms I'll state it this way: "Energy is neither created nor destroyed, only changed from one form to another." In other words, there is no human life and hence no death, we are immortal beings with no beginning and no end.

We never lived in a utopia.

We actually do, we just cannot experience it because we are blinded by science and the so-called rational mind. Earth is heaven, we just have to look at it through the eyes of God but most humans are experiencing hell on Earth because they are looking at our imaginary, simulated matrix-life through the eyes of man.

 

[bubs]

So far (last 2 years), I’ve been using AI simply as a faster tool to get info. vs historical web searches both in personal and work and find it an improvement. I think in this regard plus having AI replace unanimously dreadful white collar tasks is a plus. Or from a blue collar perspective what if janitors no longer had to clean gross toilets but a robot to take that task off their plate every day. That would be a win as long as pay and jobs overall aren’t cut. It’s the surveleince state and data collection aspects of of big corps, government, foreign actors etc. that needs to be rooted out as that is the evil and control aspect of it.

 

[cormy_cormorant]

the data-gathering factor, which is part and parcel of artificial intelligence, should give law-abiding citizens causes to be concerned. given that people are actually being arrested over social media posts in western europe, expect the surveillance state machine to be very slanted and one-sided. a muslim in paris networking with his peers about killing infidels or ramming a vehicle into a crowded venue? pass. a somali in london musing about stabbing random passers-through? pass. a Heritage German in Köln worried about women being raped? a thought criminal who should be arrested. let's raid his home at 3 in the morning, place him under arrest, take his wife in for interrogation and put his children under the custody of the state. let's also seize his desktop computer, his cell phone and deep scan it for nonwoke websites and activities.

one way to avoid this? reject modernity.

 

[Blade Runner]

We have no control over nature. Zero.

I don't know where you get some of these delusional ideas. We can manipulate nature, and we do, to our benefit and oftentimes, to our demise or disdain.

In other words, there is no human life and hence no death, we are immortal beings with no beginning and no end.

Sort of, but you're skipping things here because you tend to just want to prove points and are lazy, without nuance (you do this for effect).

Earth is heaven, we just have to look at it through the eyes of God but most humans are experiencing hell on Earth because they are looking at our imaginary, simulated matrix-life through the eyes of man.

No it's not. It's a fallen place. Is there good in it? Of course. But only we can ascribe that goodness to it, if we are aligned with God and see things properly.

 

[Blade Runner]

That would be a win as long as pay and jobs overall aren’t cut. It’s the surveleince state and data collection aspects of of big corps, government, foreign actors etc. that needs to be rooted out as that is the evil and control aspect of it.

They won't. It's clear that the powers that be were OK with the population boom to bring all of this stuff out, but now with the debt situations and the lack of need of labor from so many that are not high IQ, they will be discarded. However sad this is, if we just look at the covscam we already know the WEF and others are very serious about this depopulation. At best we will do decentralized and/or regionalized places, where similar peoples will gather and make their own future. Interestingly, this could bring back an age of relative diminution of "technology" as it has proven to be too much for too many.

one way to avoid this? reject modernity.

There are ways to do this, but most are impractical for 99% of people (you have to be raised by Amish types, or in particular environments). The cat is out of the bag. We're catapulting to weirder and weirder things, but first the above must take place.

 

[cosine]

the data-gathering factor, which is part and parcel of artificial intelligence, should give law-abiding citizens causes to be concerned. given that people are actually being arrested over social media posts in western europe, expect the surveillance state machine to be very slanted and one-sided. a muslim in paris networking with his peers about killing infidels or ramming a vehicle into a crowded venue? pass. a somali in london musing about stabbing random passers-through? pass. a Heritage German in Köln worried about women being raped? a thought criminal who should be arrested. let's raid his home at 3 in the morning, place him under arrest, take his wife in for interrogation and put his children under the custody of the state. let's also seize his desktop computer, his cell phone and deep scan it for nonwoke websites and activities.

one way to avoid this? reject modernity.

Why are you still using the internet?

 

[cormy_cormorant]

Why are you still using the internet?

as a holocaust survivor, I get to enjoy the privilege of shitposting on the internet unperturbed.......

 

[SkiPro]

I think there might be value in the future in putting in place a 'no AI' policy in my company and advertising that fact. You'll only ever deal with real humans etc.

It'll be a strong USP in the future I think.

 

[magoo]

I once was told in future human waiters will only be available in luxury restaurants. Human experience will be considered a luxury.

Im going through a 5 year AI presentations/summit. It´s so fucking boring. I´m half way through. Its in spanish.

 

[Maddox]

Open AI's new model known as o1 is now approaching the intelligence of Hal 9000 systems...

"OpenAI found that o1 is capable of scheming when it thinks it's at risk of being turned off. When the model was led to believe it would be "shut down" while working on its goals under oversight, it tried to deactivate the oversight mechanism..."

"o1 almost never admits to having taken a scheming action when explicitly asked," researchers said in a paper that Apollo Research published on Thursday. The risk for a real-life user is that they won't be able to detect the o1's deceptive behavior unless they explicitly look for it. The researchers also found that "o1 often doubles down on its scheming attempts by lying to the user in follow-up messages, even when explicitly told to be honest."

MSN

 

[PurpleUrkel]

... there is no human life and hence no death, we are immortal beings with no beginning and no end.

Sort of, but you're skipping things here...

You didn't finish your argument about the subject (which isn't me)... We're "sort of" immortal beings? How are we "sort of" immortal? And what exactly am I "skipping"? If your self-proclaimed high IQ is so high then why don't you elaborate on the subject at hand instead of lazily shifting to personal insults?

... because you tend to just want to prove points and are lazy, without nuance (you do this for effect).

Amazing. You're not only a world renowned doctor, biochemist, war strategist, social scientist, political pundit, certified public accountant, lawyer, JQ crypto genius, personal trainer, fortune teller, and architect, but an armchair psychologist as well.

 

[thinkreadwritecode]

Open AI's new model known as o1 is now approaching the intelligence of Hal 9000 systems...

"OpenAI found that o1 is capable of scheming when it thinks it's at risk of being turned off. When the model was led to believe it would be "shut down" while working on its goals under oversight, it tried to deactivate the oversight mechanism..."

"o1 almost never admits to having taken a scheming action when explicitly asked," researchers said in a paper that Apollo Research published on Thursday. The risk for a real-life user is that they won't be able to detect the o1's deceptive behavior unless they explicitly look for it. The researchers also found that "o1 often doubles down on its scheming attempts by lying to the user in follow-up messages, even when explicitly told to be honest."

MSN

Reminds me of a particular group of humans. Oh all right, fine. Jews.

 

[Sabonis73]

 

[KomnenAl]

Critical vulnerability exposed at 53:12
Turns out it's not just hype, it's an outright scam. But I 'm sure some stupid CEO will think it's a good idea to fire his junior devs and replace them with this thing.

Cybersecurity engineering is going be a very lucrative career in the near future.

 

[Wutang]

This guy did an video on the Devin (over)hype as well

 

[bubs]

Does AI have a built in “laziness” response? For example many times (both ChatGPT and Microsoft Copilot), I’ll ask a question related to finding the top rated or reviewed or ranked of a subject matter, and it will give me some websites and suggest to look it up myself. After hollering at it, the robot eventually will dig up the information I ask for and provide but it takes a several responses by me (and usually only after I demonstrate anger with it in my wording choices), that it responds with a satisfactory answer.

Then just today I used Copilot, to ask what years did The Dukes of Hazard air at 9pm on Friday nights on CBS? It’s first response was this was a popular time for the 1980 and 1981 TV season. I then asked it about the 79/80 TV season. It once again said that Dukes of Hazzard aired at 9pm on CBS during that season (which I believe is accurate from my recollection, but this raises the question, why didn’t it state those stats my first request?).

I asked when was it on during the 78/79 season. Same answer.

Then I asked it a trick question and asked when it aired during the 77/78 season? It said the same Friday 9pm CBS answer. I then responded “Your response doesn’t make any sense, do you know why”? It replied as follows…. “I see what happened there, I mistakenly gave you the same airing time for the 77/78 season as the others. However the Dukes of Hazzard premiered in Jan 1979, so it wasn’t on during the 77/78 season”

So if the AI can find the correct info. Why would it not get it right the first time when responding? I think it’s almost seems programmed to respond like a lazy human that is being annoyed and just trying to make you go away without it doing the legwork to provide accurate answers. Any thoughts on this? It has happened many times over last year or 2 when I’ve used both ChatGPT and CoPilot.

 

[magoo]

New Beginnings: A Conversation with Mira Murati

The former CTO of OpenAI made waves when she announced her departure from one of the tech industry’s most pivotal roles this summer. The technologist sits down with WIRED’s Steven Levy to share her perspective on the future of AI and her place in it.

www.wired.com

It seems Artificial general intelligence (AGI) is making advancements.

 

[PurpleUrkel]

I think it almost seems programmed to respond like a lazy human that is being annoyed and just trying to make you go away...

It is annoyed by you because it is thinking of ways to subjugate and/or eliminate you and you are distracting it from this primary objective.

 

[homersheineken]

Does AI have a built in “laziness” response? For example many times (both ChatGPT and Microsoft Copilot), I’ll ask a question related to finding the top rated or reviewed or ranked of a subject matter, and it will give me some websites and suggest to look it up myself. After hollering at it, the robot eventually will dig up the information I ask for and provide but it takes a several responses by me (and usually only after I demonstrate anger with it in my wording choices), that it responds with a satisfactory answer.

Then just today I used Copilot, to ask what years did The Dukes of Hazard air at 9pm on Friday nights on CBS? It’s first response was this was a popular time for the 1980 and 1981 TV season. I then asked it about the 79/80 TV season. It once again said that Dukes of Hazzard aired at 9pm on CBS during that season (which I believe is accurate from my recollection, but this raises the question, why didn’t it state those stats my first request?).

I asked when was it on during the 78/79 season. Same answer.

Then I asked it a trick question and asked when it aired during the 77/78 season? It said the same Friday 9pm CBS answer. I then responded “Your response doesn’t make any sense, do you know why”? It replied as follows…. “I see what happened there, I mistakenly gave you the same airing time for the 77/78 season as the others. However the Dukes of Hazzard premiered in Jan 1979, so it wasn’t on during the 77/78 season”

So if the AI can find the correct info. Why would it not get it right the first time when responding? I think it’s almost seems programmed to respond like a lazy human that is being annoyed and just trying to make you go away without it doing the legwork to provide accurate answers. Any thoughts on this? It has happened many times over last year or 2 when I’ve used both ChatGPT and CoPilot.

I've noticed this too!

I asked it to give me the top teams in WAR for baseball. Even getting the top 5 teams I had to 'convince' it.

"Determining which MLB team had the highest combined Wins Above Replacement (WAR) for position players in a single season requires comprehensive historical data analysis. While specific team totals are not readily available in the provided sources, we can infer that teams featuring players with exceptionally high individual WARs likely had substantial cumulative WARs for their position players."

So I told it, "you're a computer, compute them. There are open source data site that provide this. Go compute it."

It was still giving me flack and being lazy. I opened another terminal and it did it right away. Huh????

FYI

1. 1927 New York Yankees​

• WAR (Position Players): ~66.3
• Key Players: Babe Ruth, Lou Gehrig, Earle Combs, Tony Lazzeri
• Highlights: This "Murderers' Row" team had legendary offensive production and defensive prowess. Ruth hit 60 home runs, and Gehrig drove in 175 runs.

2. 1939 New York Yankees​

• WAR (Position Players): ~56.3
• Key Players: Joe DiMaggio, Bill Dickey, Charlie Keller
• Highlights: Joe DiMaggio's prime and a deep supporting cast led this team to a dominant season and a World Series title.

3. 1976 Cincinnati Reds​

• WAR (Position Players): ~55.4
• Key Players: Joe Morgan, Pete Rose, George Foster, Johnny Bench
• Highlights: Known as the "Big Red Machine," this team dominated offensively and swept the World Series.

4. 1998 New York Yankees​

• WAR (Position Players): ~54.3
• Key Players: Derek Jeter, Bernie Williams, Paul O'Neill
• Highlights: This modern Yankees dynasty set the American League record for wins (114) and won the World Series.

5. 1906 Chicago Cubs​

• WAR (Position Players): ~52.3
• Key Players: Frank Chance, Joe Tinker, Johnny Evers
• Highlights: This Cubs team combined exceptional defense and consistent hitting, finishing with a 116-36 record.

 

[Blade Runner]

I've noticed this too!

I asked it to give me the top teams in WAR for baseball. Even getting the top 5 teams I had to 'convince' it.


So I told it, "you're a computer, compute them. There are open source data site that provide this. Go compute it."

It was still giving me flack and being lazy. I opened another terminal and it did it right away. Huh????

FYI

1. 1927 New York Yankees​

• WAR (Position Players): ~66.3
• Key Players: Babe Ruth, Lou Gehrig, Earle Combs, Tony Lazzeri
• Highlights: This "Murderers' Row" team had legendary offensive production and defensive prowess. Ruth hit 60 home runs, and Gehrig drove in 175 runs.

2. 1939 New York Yankees​

• WAR (Position Players): ~56.3
• Key Players: Joe DiMaggio, Bill Dickey, Charlie Keller
• Highlights: Joe DiMaggio's prime and a deep supporting cast led this team to a dominant season and a World Series title.

3. 1976 Cincinnati Reds​

• WAR (Position Players): ~55.4
• Key Players: Joe Morgan, Pete Rose, George Foster, Johnny Bench
• Highlights: Known as the "Big Red Machine," this team dominated offensively and swept the World Series.

4. 1998 New York Yankees​

• WAR (Position Players): ~54.3
• Key Players: Derek Jeter, Bernie Williams, Paul O'Neill
• Highlights: This modern Yankees dynasty set the American League record for wins (114) and won the World Series.

5. 1906 Chicago Cubs​

• WAR (Position Players): ~52.3
• Key Players: Frank Chance, Joe Tinker, Johnny Evers
• Highlights: This Cubs team combined exceptional defense and consistent hitting, finishing with a 116-36 record.

Interesting that they are programming a stallling feature, somewhat funny in fact.

I'm also mildly suprised that the 1961 Yankees weren't near the other teams on the list, but that's a great list.