AI - The artificial intelligence thread

[usagi]

22 hours ago, ignatius said:

aka boundary testing, something any software team worth its salt would do anyway.

you're welcome that the public at large is doing your work for you.

[usagi]

[auxien]

17 hours ago, Alcofribas said:

otoh I find it frustrating how incapable we seem to be able to wrestle with this as a public.

honestly it's pretty well expected. the general public doesn't have much interest in such things, sadly. we're trained by Western societal design to not. see "our society is dominated by a craven political class, techno hucksters, and a media that is hardly more than a marketing department for those who use the system to dominate and hoard wealth." they don't want us thinking or talking or trying past the basic boring things that allow for society to keep churning.

17 hours ago, Alcofribas said:

I’d like to be hopeful with you that we’ll be forced to grow to a greater understanding of human nature or something but it’s hard not to be cynical.

i have plenty of both in me. i do not recommend this route my good sir Alco, but it's possible. cynical beyond reason at near everything, but hopeful that i'm often wrong to be so cyncial.

17 hours ago, Alcofribas said:

we still have oppressed people struggling for basic rights like trans people in the US, why would some computer be the thing that finally gets us to dig a little deeper? look also at the grotesque ways people treat other intelligent life forms that occupy this planet - with indifference, torture, exploitation.

you're not wrong, but it seems to me that US society is always dragging the oppressed along as it 'advances' ...it's not a 3 steps forward, 2 steps back thing but more of a 1 slight shuffle ahead, while the lagging leg drags farther behind? a bit of a tortured metaphor...but i hope it makes a little sense. anyway: i don't think deeper understanding of human consciousness and intelligence will fix...well, anything. it won't bring acceptance or equal rights for trans persons or any other oppressed groups.

deeper understanding of our intelligence may give us some better perspective on intelligence all around us, our place in the world with more respect all around, slightly...maybe. a little. any step in that direction is likely to do some good in the long run tho. this is a hopeful thought tho...it could have an opposite effect of triggering kneejerk reactions of deeper conservatism, rejection of 'progress' etc. very, very hard to say what will or even could happen.

18 hours ago, Alcofribas said:

i genuinely struggle to see AI as anything but a tool of the powerful to further expand their brutal rule and enrich themselves. the people will get scraps to help them organize the files on the computer or write their essays or whatever. 

there's a good likelihood that any technological progress will bring exactly what you're expecting, history tends to suggest this...in the short term. industrial revolution brought about horrid working conditions and a push for more workers to keep in the factories, ages or education or their health be damned...but in the long term the industrial revolution & its further advancements have lead to changes that are by and large better for much of Western society (at least by many metrics). there's no guarantee of this with AI/AGI/etc. or the effects it may have on our understanding of intelligence, consciousness, etc. i'm not saying we should be looking forward to or encouraging these people, thinking 'well if they wreck our current livelihoods maybe it'll turn around eventually' ...surely not. i'm just looking at likely paths in the large scale, on the long term.

i guess my point is any technology is a likely to be a tool of the powerful to expand their brutal rule and enrich themselves. one can accept this or not, but history makes this pretty clear from the little i know of it. if you want to rail against this one or any other new legit/viable tech, i get it, but it's a battle you won't win. i think the only solution is to try and level the playing field as much as possible, and i see things like deeper knowledge, understanding, etc. as the only true way to start down that path.

[luke viia]

hey I feel like rambling on a topic I know very little about, so suffer me now:

ahem. even if AI/AGI gives "us" a better understanding of consciousness/our place in the world etc, it's a little odd to think that an intellectual advance like that will have an existentially meaningful impact on most people's day to day lives. we'll still live in the world in our own little monads, and AI isn't going to just override the daily experience/confusion of living in the world. our understanding of ourselves is a very personal affair, it isn't a simultaneous species-wide advancement, and I'm certain that technology cannot tell us who or what we actually are on the most important levels, nor can the broader category of technological sciences. even if AI causes "us" to stop and reflect on our human situation at some point, it won't alter our basic feeling/apprehension of ourselves in any sustained way - that is something that only occurs person by person, and is rewarded directly by how seriously one takes the problem and investigates it for themselves. we are not all on the same level. AI can take over the world and there will still be mouth-breathing couch potatoes and deeply unbalanced individuals marching alongside the scholars and intelligentsia. any and all talk of "bettering humanity" through technology like this is nonsense, it's just this decade's "better living through chemistry." we all have to do the work ourselves; all the world's knowledge at our collective fingertips won't do a thing to change our course if we don't internalize it - and that's what worries me about AI. this insane promise that having all this knowledge will somehow make us better, by pure access alone. it's ridiculous. merely owning an encyclopedia doesn't make a person more intelligent

okay now somebody ask chatgpt to summarize my post. looking at you, ooooo

[vkxwz]

6 minutes ago, luke viia said:

hey I feel like rambling on a topic I know very little about, so suffer me now:

ahem. even if AI/AGI gives "us" a better understanding of consciousness/our place in the world etc, it's a little odd to think that an intellectual advance like that will have an existentially meaningful impact on most people's day to day lives. we'll still live in the world in our own little monads, and AI isn't going to just override the daily experience/confusion of living in the world. our understanding of ourselves is a very personal affair, it isn't a simultaneous species-wide advancement, and I'm certain that technology cannot tell us who or what we actually are on the most important levels, nor can the broader category of technological sciences. even if AI causes "us" to stop and reflect on our human situation at some point, it won't alter our basic feeling/apprehension of ourselves in any sustained way - that is something that only occurs person by person, and is rewarded directly by how seriously one takes the problem and investigates it for themselves. we are not all on the same level. AI can take over the world and there will still be mouth-breathing couch potatoes and deeply unbalanced individuals marching alongside the scholars and intelligentsia. any and all talk of "bettering humanity" through technology like this is nonsense, it's just this decade's "better living through chemistry." we all have to do the work ourselves; all the world's knowledge at our collective fingertips won't do a thing to change our course if we don't internalize it - and that's what worries me about AI. this insane promise that having all this knowledge will somehow make us better, by pure access alone. it's ridiculous. merely owning an encyclopedia doesn't make a person more intelligent

okay now somebody ask chatgpt to summarize my post. looking at you, ooooo

Yeah I don't see it being likely that creating agi will give the layperson a better understanding of themselves, and even if as a society we understand how the brain itself functions in a general level, it'll still be, as you said, a very personal affair for someone to go on the journey of understanding their own specific mind. I suspect we'd get some pretty interesting insights but the average person probably won't care anyway. I do think that the promise of AI isn't just information though, it's the extra productivity without the need for as many people.

[usagi]

7 hours ago, luke viia said:

hey I feel like rambling on a topic I know very little about, so suffer me now:

ahem. even if AI/AGI gives "us" a better understanding of consciousness/our place in the world etc, it's a little odd to think that an intellectual advance like that will have an existentially meaningful impact on most people's day to day lives. we'll still live in the world in our own little monads, and AI isn't going to just override the daily experience/confusion of living in the world. our understanding of ourselves is a very personal affair, it isn't a simultaneous species-wide advancement, and I'm certain that technology cannot tell us who or what we actually are on the most important levels, nor can the broader category of technological sciences. even if AI causes "us" to stop and reflect on our human situation at some point, it won't alter our basic feeling/apprehension of ourselves in any sustained way - that is something that only occurs person by person, and is rewarded directly by how seriously one takes the problem and investigates it for themselves. we are not all on the same level. AI can take over the world and there will still be mouth-breathing couch potatoes and deeply unbalanced individuals marching alongside the scholars and intelligentsia. any and all talk of "bettering humanity" through technology like this is nonsense, it's just this decade's "better living through chemistry." we all have to do the work ourselves; all the world's knowledge at our collective fingertips won't do a thing to change our course if we don't internalize it - and that's what worries me about AI. this insane promise that having all this knowledge will somehow make us better, by pure access alone. it's ridiculous. merely owning an encyclopedia doesn't make a person more intelligent

okay now somebody ask chatgpt to summarize my post. looking at you, ooooo

I mostly agree with this. I think there may be some insights to be gained from modeling human consciousness in an artificial replica and playing around with that model. however my concern is with the practice/application, not the experiments. same as most skeptics of this technology, I worry this will be used to accelerate the growing divide between the haves and the have-nots. the classic techbro mistake is thinking that everyone falls into the same basic character profile when the majority of the people on this planet are completely different people living lives that are not within the lived understanding of the average westerner. those people will be left behind. or worse, if they have the capacity to, they will build competing systems which, instead of uniting harmoniously with the rest for the "benefit of all mankind", will simply become the next battlespace or arms race.

related to that last point, I read the other day that Altman's a prepper too. he's bragged about having his own plot of land in Big Sur and stocks to live through nuclear destruction. the more you find out about this guy the more there is to distrust.

[ignatius]

a nice walk through the "Big Data" hype era into Ai. nuts and bolts of the bullshit.

and typical tech internet youtube takedown style video about the R1 and the guy behind it being full of shit. not surprising. 

edit: the microsoft copilot hate videos are are filling youtube algorithms

 

Edited May 30 by ignatius

[Rubin Farr]

[Nebraska]

On 5/24/2024 at 1:31 PM, logakght said:

Yeah, fuck Sam Altman—now I completely understand Ilya's departure.

 

On 5/24/2024 at 1:34 PM, Rubin Farr said:

He seems to flat out lie a lot.

great, he's now part of openAI's oversight team now

[zero]

more insight into the Sam Altman drama:

https://www.cnbc.com/2024/05/29/former-openai-board-member-explains-why-ceo-sam-altman-was-fired.html

he definitely sounds like a toxic guy. probably not the type of person that should be leading the charge toward "humanity's future" or whatever it's called...wonder who will play him in the Hulu mini-series that will surely come out some day about all this. Paul Reiser's way too old now

[Nebraska]

6 minutes ago, zero said:

more insight into the Sam Altman drama:

https://www.cnbc.com/2024/05/29/former-openai-board-member-explains-why-ceo-sam-altman-was-fired.html

he definitely sounds like a toxic guy. probably not the type of person that should be leading the charge toward "humanity's future" or whatever it's called...wonder who will play him in the Hulu mini-series that will surely come out some day about all this. Paul Reiser's way too old now

don't worry, he's promised to donate at least a billion dollars to charity so "it's cool"

[logakght]

5 minutes ago, Nebraska said:

don't worry, he's promised to donate at least a billion dollars to charity so "it's cool"

Is he an "Effective Altruist" too? I believe this movement is a complete scam.

[auxien]

38 minutes ago, Nebraska said:

don't worry, he's promised to donate at least a billion dollars to charity so "it's cool"

lol wow it’s hard to be surprised with how fucking boring, pathetic, & 100% transparent these shitfucks are, and yet i am. 

[Nebraska]

30 minutes ago, logakght said:

Is he an "Effective Altruist" too? I believe this movement is a complete scam.

this is:

Created by Warren Buffett, Melinda French Gates, and Bill Gates, the Giving Pledge came to life following a series of conversations with philanthropists about how they could set a new standard of generosity among the ultra-wealthy.

The Giving Pledge is a simple concept: an open invitation for billionaires, or those who would be if not for their giving, to publicly commit to give the majority of their wealth to philanthropy either during their lifetimes or in their wills.

not sure how "generous" the others have been, but melinda gates is definitely doing her thing

[zero]

1 hour ago, Nebraska said:

don't worry, he's promised to donate at least a billion dollars to charity 

in like a 100 word or less blurb, I noticed the word "scaffolding" is used 3 times. total chat gpt hack job. mofo can't even write anything heartfelt. 

[Nebraska]

56 minutes ago, zero said:

in like a 100 word or less blurb, I noticed the word "scaffolding" is used 3 times. total chat gpt hack job. mofo can't even write anything heartfelt. 

also hope they fix this. maybe get chatGPT to proof-read work before replying.

can't be caught lacking like this out here

[Rubin Farr]

 

https://m.youtube.com/watch?v=_9mces4HKNM

[o00o]

this rabbit thing is just so scammy

[Nebraska]

On 5/25/2024 at 8:44 AM, ignatius said:

Google claims some widely distributed screenshots of AI Overviews gone wrong were fake, which seems to be true based on WIRED’s own testing. For example, a user on X posted a screenshot that appeared to be an AI Overview responding to the question “Can a cockroach live in your penis?” with an enthusiastic confirmation from the search engine that this is normal. The post has been viewed over 5 million times. Upon further inspection, though, the format of the screenshot doesn’t align with how AI Overviews are actually presented to users.

Liz Reid (head of search @ Google) post directly referenced two of the most viral, and wildly incorrect, AI Overview results.

One saw Google's algorithms endorse eating rocks because doing so “can be good for you,” and the other suggested using nontoxic glue to thicken pizza sauce.

Rock eating is not a topic many people were ever writing or asking questions about online, so there aren't many sources for a search engine to draw on. According to Reid, the AI tool found an article from The Onion, a satirical website, that had been reposted by a software company, and it misinterpreted the information as factual. As for Google telling its users to put glue on pizza, Reid effectively attributed the error to a sense of humor failure.

“We saw AI Overviews that featured sarcastic or troll-y content from discussion forums,” she wrote. “Forums are often a great source of authentic, first-hand information, but in some cases can lead to less-than-helpful advice, like using glue to get cheese to stick to pizza.”

https://www.wired.com/story/google-ai-overview-search-issues/

[Alcofribas]

"misinterpreted the information as factual," "a sense of humor failure."

no

 

[usagi]

1 hour ago, Nebraska said:

For example, a user on X posted a screenshot that appeared to be an AI Overview responding to the question “Can a cockroach live in your penis?” with an enthusiastic confirmation from the search engine that this is normal.

this one was obviously fake and was posted around for the lols. however if someone sees a screenshot like that and accepts it at face value as an example of dumb AI in effect, that still highlights the same problem.

[Rubin Farr]

7 hours ago, Nebraska said:

Google claims some widely distributed screenshots of AI Overviews gone wrong were fake, which seems to be true based on WIRED’s own testing. For example, a user on X posted a screenshot that appeared to be an AI Overview responding to the question “Can a cockroach live in your penis?” with an enthusiastic confirmation from the search engine that this is normal. The post has been viewed over 5 million times. Upon further inspection, though, the format of the screenshot doesn’t align with how AI Overviews are actually presented to users.

Liz Reid (head of search @ Google) post directly referenced two of the most viral, and wildly incorrect, AI Overview results.

One saw Google's algorithms endorse eating rocks because doing so “can be good for you,” and the other suggested using nontoxic glue to thicken pizza sauce.

Rock eating is not a topic many people were ever writing or asking questions about online, so there aren't many sources for a search engine to draw on. According to Reid, the AI tool found an article from The Onion, a satirical website, that had been reposted by a software company, and it misinterpreted the information as factual. As for Google telling its users to put glue on pizza, Reid effectively attributed the error to a sense of humor failure.

“We saw AI Overviews that featured sarcastic or troll-y content from discussion forums,” she wrote. “Forums are often a great source of authentic, first-hand information, but in some cases can lead to less-than-helpful advice, like using glue to get cheese to stick to pizza.”

https://www.wired.com/story/google-ai-overview-search-issues/

 

[ignatius]

20 hours ago, Rubin Farr said:

 

that's kinda pretty funny. feedback of garbage from memes and jokes. it eats its own tail. 

Ai generated articles.. 

 

[Rubin Farr]

 

[auxien]

31 minutes ago, Rubin Farr said:

 

"here at the world's best lava gun company, we wanted to let you know that some buyers of our cheap and cutting edge product, available now at the link below, may be using the lava gun to enact harm upon others by shooting lava at them. this nefarious action was unimaginable by us, and is obviously the fault of the evil actors buying them which is entirely against the terms and conditions, and we hope you stay at least 2 lava-shot distances away from them.

we'd like to take this opportunity to mention an exciting new update to the lava gun which triples the range of lava shot is being pushed out tonight. sign up tonight to be the first in your neighborhood to try it out!"