AI - The artificial intelligence thread

this topic reminds me of crypto/bitcoin/nft quite a lot. like, we have this crypto thread on watmm where there was this core idea about how crypto was like some big new step for "humanity" or computing or whatever. several years down the line, does that appear to have any resemblance to reality? definitely not, to me. there's this core of bullshit to all this and it reeks of a scam that keeps getting repeated over and over again. extremely rich tech giants produce products (actually, it's just as common to have no product whatsoever) and everything about it is discussed as this big revolution for humanity. it's totally the future, it's inevitable, this changes everything, there's no going back. the media does what it does best, i.e., transcribes the marketing copy with total credulity. social media is inundated with all incessant content about some superficial aspect of this stuff - remember when we were seeing "apes" over and over for however long that was? some dipshits created nft apes and we just had to look at this everyday like this is important in any way whatsoever. in the end, what gets delivered? what problem is ai art "solving?" what is the point of gpt? i've yet to see anything even remotely like some kind of revolutionary power here. i'm seeing an incredibly inefficient technology that requires massive computing power to produce a vague product that con artists like sam fucking altman get rich talking about like cartoon salesman villains from an 80s horror satire movie. i'm tired of it. i'm tired of people like elon musk and sam altman talking about how what they do is good for humanity and establishment media uncritically reporting all this as anything other than schemes to get rich by hooking the public on stupid technology. i mean, look at this recent brain dead chicanery by altman just ripping off scarlett johanson's voice with his idiot technology* and then pretending it wasn't on purpose or whatever. these guys can't help themselves, they have to shove their little dicks around and they're rich enough that they can be transparently full of shit and get away with it over and over. it's pathetic. 

otherwise, yeah it's really cool. 

*it's worth thinking about this one bc altman tweeted "her" in obvious reference to the scifi movie about ai. this is the incredible inevitable future we are being promised? a stupid ripoff of a spike jonze movie from 2013? and the tech bozo doesn't even pay the voice actor lol. come on man

  On 5/22/2024 at 9:10 PM, Alcofribas said:

what gets delivered? what problem is ai art "solving?" what is the point of gpt?

thats actually the core issue. currently its quite a lot

  On 5/22/2024 at 9:10 PM, Alcofribas said:

this topic reminds me of crypto/bitcoin/nft quite a lot. like, we have this crypto thread on watmm where there was this core idea about how crypto was like some big new step for "humanity" or computing or whatever. several years down the line, does that appear to have any resemblance to reality? definitely not, to me. there's this core of bullshit to all this and it reeks of a scam that keeps getting repeated over and over again. extremely rich tech giants produce products (actually, it's just as common to have no product whatsoever) and everything about it is discussed as this big revolution for humanity. it's totally the future, it's inevitable, this changes everything, there's no going back. the media does what it does best, i.e., transcribes the marketing copy with total credulity. social media is inundated with all incessant content about some superficial aspect of this stuff - remember when we were seeing "apes" over and over for however long that was? some dipshits created nft apes and we just had to look at this everyday like this is important in any way whatsoever. in the end, what gets delivered? what problem is ai art "solving?" what is the point of gpt? i've yet to see anything even remotely like some kind of revolutionary power here. i'm seeing an incredibly inefficient technology that requires massive computing power to produce a vague product that con artists like sam fucking altman get rich talking about like cartoon salesman villains from an 80s horror satire movie. i'm tired of it. i'm tired of people like elon musk and sam altman talking about how what they do is good for humanity and establishment media uncritically reporting all this as anything other than schemes to get rich by hooking the public on stupid technology. i mean, look at this recent brain dead chicanery by altman just ripping off scarlett johanson's voice with his idiot technology* and then pretending it wasn't on purpose or whatever. these guys can't help themselves, they have to shove their little dicks around and they're rich enough that they can be transparently full of shit and get away with it over and over. it's pathetic. 

otherwise, yeah it's really cool. 

 

*it's worth thinking about this one bc altman tweeted "her" in obvious reference to the scifi movie about ai. this is the incredible inevitable future we are being promised? a stupid ripoff of a spike jonze movie from 2013? and the tech bozo doesn't even pay the voice actor lol. come on man

Expand  

It has changed a lot in how college students complete assignments.  My wife teaches in an Engineering dept and they have to come up with all sorts of new tools and ways to detect whether or not students are actually writing the assignments themselves or they just asked some LLM to write their assignment or write their code.

I get that it will probably become the norm even for people who write or code for a living, but I guess I'm just having an old man get off my lawn moment when I think that the students should still be able to learn to write / code on their own, regardless of whether or not an LLM can do it for them.

  On 5/22/2024 at 9:10 PM, Alcofribas said:

this topic reminds me of crypto/bitcoin/nft quite a lot. like, we have this crypto thread on watmm where there was this core idea about how crypto was like some big new step for "humanity" or computing or whatever. several years down the line, does that appear to have any resemblance to reality? definitely not, to me. there's this core of bullshit to all this and it reeks of a scam that keeps getting repeated over and over again. extremely rich tech giants produce products (actually, it's just as common to have no product whatsoever) and everything about it is discussed as this big revolution for humanity. it's totally the future, it's inevitable, this changes everything, there's no going back. the media does what it does best, i.e., transcribes the marketing copy with total credulity. social media is inundated with all incessant content about some superficial aspect of this stuff - remember when we were seeing "apes" over and over for however long that was? some dipshits created nft apes and we just had to look at this everyday like this is important in any way whatsoever. in the end, what gets delivered? what problem is ai art "solving?" what is the point of gpt? i've yet to see anything even remotely like some kind of revolutionary power here. i'm seeing an incredibly inefficient technology that requires massive computing power to produce a vague product that con artists like sam fucking altman get rich talking about like cartoon salesman villains from an 80s horror satire movie. i'm tired of it. i'm tired of people like elon musk and sam altman talking about how what they do is good for humanity and establishment media uncritically reporting all this as anything other than schemes to get rich by hooking the public on stupid technology. i mean, look at this recent brain dead chicanery by altman just ripping off scarlett johanson's voice with his idiot technology* and then pretending it wasn't on purpose or whatever. these guys can't help themselves, they have to shove their little dicks around and they're rich enough that they can be transparently full of shit and get away with it over and over. it's pathetic. 

otherwise, yeah it's really cool. 

 

*it's worth thinking about this one bc altman tweeted "her" in obvious reference to the scifi movie about ai. this is the incredible inevitable future we are being promised? a stupid ripoff of a spike jonze movie from 2013? and the tech bozo doesn't even pay the voice actor lol. come on man

Expand  

Yeah I can see how if you're just doing pattern matching you go: big hype = not panning out in the future, and so you see the big hype and go, well I've seen this before, it's going to be nothing again. In this case if you engage with the actual topic, thinking it will turn out to be nothing means that either you believe we won't get anything close to human level intelligence in silicon any time soon, or you believe we will but that it won't matter. The former position is valid, but all the hype is coming from extrapolating from the very real progress that exists in that field. Nobody sane is saying chatgpt in it's current state will totally revolutionize the world (although I know quite a few people that use it on a nearly daily basis to save a lot of time), but they are extrapolating from the progress in deep learning in general. It was only 12 years ago that deep learning got 16% accuracy on imagenet (image classification tasks), by 2017 most of the best systems had over 95% accuracy. In 2018 gpt 1 was trained and was hopeless, gpt 2 and 3 were huge improvements and gpt4 is now doing things that experts in the field a few years ago would have said was 20 years away or more. So if we end up with nothing happening, it'll be because deep learning hits a wall that there isn't evidence of yet, or because having human level intelligence in silicon will have no impact. And if you're claiming the former, the current evidence is against you.

  On 5/22/2024 at 9:19 PM, EdamAnchorman said:

I get that it will probably become the norm even for people who write or code for a living, but I guess I'm just having an old man get off my lawn moment when I think that the students should still be able to learn to write / code on their own, regardless of whether or not an LLM can do it for them.

Yeah I agree with you, I think learning to write is super useful and if we lose that people will be much worse off. And surely it's not a super hard problem to fix, just have more writing exams that are done in class where you don't see the question beforehand.

  On 5/16/2024 at 12:49 AM, jaderpansen said:

 

Expand  

she's advertising AI products in this very video, wtf?

  On 5/16/2024 at 5:14 AM, auxien said:

Ilya was one of the ones privately pushing for Altman's ouster a few months ago if my memory serves me correctly. curious what he's going to do that Altman is describing as 'focused on getting to the best future for humanity' and how that's different than this Jakub who's going to 'lead us to make rapid and safe progress towards our mission of ensuring that AGI benefits everyone' - i'm gathering these two cats were in opposition to each other in some sense or another... the 'safe' in there is obviously overcompensating for the expected 'unsafe' roads they're definitely going to actually take.

Expand  

via theverge.com

  On 5/23/2024 at 3:11 AM, usagi said:

 

Expand  

the host of this plays the part perfectly. good one liners throughout. 

  On 5/22/2024 at 9:10 PM, Alcofribas said:

this topic reminds me of crypto/bitcoin/nft quite a lot. like, we have this crypto thread on watmm where there was this core idea about how crypto was like some big new step for "humanity" or computing or whatever. several years down the line, does that appear to have any resemblance to reality? definitely not, to me. there's this core of bullshit to all this and it reeks of a scam that keeps getting repeated over and over again. extremely rich tech giants produce products (actually, it's just as common to have no product whatsoever) and everything about it is discussed as this big revolution for humanity. it's totally the future, it's inevitable, this changes everything, there's no going back. the media does what it does best, i.e., transcribes the marketing copy with total credulity. social media is inundated with all incessant content about some superficial aspect of this stuff - remember when we were seeing "apes" over and over for however long that was? some dipshits created nft apes and we just had to look at this everyday like this is important in any way whatsoever. in the end, what gets delivered? what problem is ai art "solving?" what is the point of gpt? i've yet to see anything even remotely like some kind of revolutionary power here. i'm seeing an incredibly inefficient technology that requires massive computing power to produce a vague product that con artists like sam fucking altman get rich talking about like cartoon salesman villains from an 80s horror satire movie. i'm tired of it. i'm tired of people like elon musk and sam altman talking about how what they do is good for humanity and establishment media uncritically reporting all this as anything other than schemes to get rich by hooking the public on stupid technology. i mean, look at this recent brain dead chicanery by altman just ripping off scarlett johanson's voice with his idiot technology* and then pretending it wasn't on purpose or whatever. these guys can't help themselves, they have to shove their little dicks around and they're rich enough that they can be transparently full of shit and get away with it over and over. it's pathetic. 

otherwise, yeah it's really cool. 

 

*it's worth thinking about this one bc altman tweeted "her" in obvious reference to the scifi movie about ai. this is the incredible inevitable future we are being promised? a stupid ripoff of a spike jonze movie from 2013? and the tech bozo doesn't even pay the voice actor lol. come on man

Expand  

nfts/bitcoin/etc are hype bullshit, 100%. but the underlying tech will at some point be a viable component (or predecessor to an eventual solution) for 'digital' cash. execution = bad, underlying ideas = maybe has a place.

AI/ML has huge revolutionary implications for all kinds of data sorting, in good ways (science, health tracking assessments, etc.) and also for all sorts of stupid/bad/horrid shit (writing college essays, tracking users online, etc.). it's a valid tool... however, ChatGPT...maybe it has some uses (see o00o's comments, already being used for coding by some, etc.) but i can't say i'm interested in any particularly.

there's a writer whose newsletter i read sometimes, he's documented some benefits (not ChatGPT specifically but some program or programs built with OpenAI or similar 'AI' tools) at sorting his ideas for stories/novels, trying to suggest multiple connections between characters, things like that. i'm not a writer but that sounds like the fun part to me, but he's spoken of it being a help in suggesting things he'd not considered, and tracking plot details in the story threads, shit like that... (how many times did Alice refer to weapons or guns in the first 3 chapters? that sorta thing i think could make sense for an AI tool as part of research, even if i don't think i'd suggest it necessarily)... the tech will at some point likely be good enough to replace some actual customer service type inquiries (taking away human jobs), there's any number of other potentials....but it could largely fizzle out too outside of some niche applications (it already has a guaranteed future in the sciences from everything i'm gathering)

hopefully it will help sort human 'work' to humans, and allow for additional tools for those humans at times, but also start to shift the workforce away from some particularly menial/repetitive job types. but i'm sure as shit no expert...

this is top of mind, just watched it yesterday...in the last section here the director of Astrophysics at NASA is talking about trying to figure out how exactly they're going to deal with all the terabytes of data coming from the current and future missions etc....making it available to all and, key here, *not having to download all of it just to work with it by individual persons* ...he doesn't say AI/ML tools, but, i'm assuming that's gotta be what they're going with. i can't imagine what other kinds of tools could do these things (but again, i'm no expert so maybe i'm assuming here). nonetheless others working in astrophysics/etc. are using and will be using tools like this more and more.

the inefficient tech is horrid. ai art/music/writing is fucking stupid, tho sometimes impressive, it's beyond mastubatory. by and large 'creations' of this tech is bad. using the tech as a tool in part of a valid reasoning tho, that has some point (again, not necessarily efficient or 'good' but, it makes some sense). it's probably been said in this thread a half dozen times but someone said something like 'why should i listen to music that someone couldn't be bothered to write and record' and ...yeah, that sums any 'creation' AI imo.

tl;dr yeah it's largely grifting rn, but there's valid useful partly revolutionary tech underlying it which is already being used for 'good' & 'useful' purposes.

  On 5/23/2024 at 1:48 AM, vkxwz said:

Yeah I can see how if you're just doing pattern matching you go: big hype = not panning out in the future, and so you see the big hype and go, well I've seen this before, it's going to be nothing again. In this case if you engage with the actual topic, thinking it will turn out to be nothing means that either you believe we won't get anything close to human level intelligence in silicon any time soon, or you believe we will but that it won't matter.

or:

AI that outperforms humans in a working capacity (i.e. just in terms of capability and responsiveness before you even consider the abstract notion of "intelligence") might well be possible soon, and it is likely to be a net bad for mankind. in the same way that every other touted tech miracle has been a net bad for us, by further entrenching the worst of our tendencies and excesses and creating new, unforeseen problems. "but it solved some of the old problems!" the silicon valley slumlords and rentseekers will splutter, which is not a worthwhile counterargument.

in any case, I don't think the general recurring pattern of scientific discovery being exploited by shitty people to create self-serving anti-human systems is going to stop anytime soon. I also don't think it necessarily negates the value of discovery itself. it's not that progress in this field isn't worthwhile or exciting. this is not a simple "technology bad" argument. the concern is how it will pan out practically, for everyone, not just the advantaged - something that techbros infamously do not give a shit about, or pretend to give a shit about at best. "move fast and break things" has now been replaced by... "the best future for humanity" or whatever else Altman tweets while the people at OpenAI who might keep him in check just quit.

  On 5/23/2024 at 5:59 AM, usagi said:

or:

AI that outperforms humans in a working capacity (i.e. just in terms of capability and responsiveness before you even consider the abstract notion of "intelligence") might well be possible soon, and it is likely to be a net bad for mankind. in the same way that every other touted tech miracle has been a net bad for us, by further entrenching the worst of our tendencies and excesses and creating new, unforeseen problems. "but it solved some of the old problems!" the silicon valley slumlords and rentseekers will splutter, which is not a worthwhile counterargument.

in any case, I don't think the general recurring pattern of scientific discovery being exploited by shitty people to create self-serving anti-human systems is going to stop anytime soon. I also don't think it necessarily negates the value of discovery itself. it's not that progress in this field isn't worthwhile or exciting. this is not a simple "technology bad" argument. the concern is how it will pan out practically, for everyone, not just the advantaged - something that techbros infamously do not give a shit about, or pretend to give a shit about at best. "move fast and break things" has now been replaced by... "the best future for humanity" or whatever else Altman tweets while the people at OpenAI who might keep him in check just quit.

Expand  

Yeah I was just arguing against the "it's nothing, all hype no effect" type of take. I am in agreement that there is a lot of potential for negative impacts, and if this thing takes off quick enough I think it's quite a scary thing. I don't agree that every technological advancement has been a net negative for humanity though, I'd much rather be alive today than any other time in history. I don't want to live in fear of starvation, death from attacks by wild animals or illness, and I don't want child mortality to be ridiculously high or for child birth to be much more dangerous than it is now, and I actually prefer our hugely extended life expectancy and the massive access to information I have through the Internet and the ability to connect with many more people.

  On 5/23/2024 at 7:04 AM, vkxwz said:

I don't agree that every technological advancement has been a net negative for humanity though, I'd much rather be alive today than any other time in history. 

I meant every touted tech miracle since the age of Big Tech, particularly from the mid-2000s onwards. even that doesn't mean every technological advancement since then, it specifically means the tech/IT industries and the "this new app or platform or gadget will solve everything" mentality.

I do still think some earlier technological innovations also have pitfalls, e.g. the internet itself, which has had big positives but also continues to rewire our monkey brains and alter society in unexpected and not entirely good ways. on the whole though, these were driven by better guiding principles and more oversight. with the wild west growth of tech giants and their centralisation of power and wealth, as well as all these random-ass privately funded startups that pride themselves on "disrupting" society (i.e. finding new ways to commoditise and control everything), they have weaponised this disingenuous approach to technological progress where they pitch it as a public good while using it to serve private interests. a lot of them believe their own bullshit too because of the overpowering allure of technical sweetness. I think all this is at odds with the best interests of humanity.

Ai-free search results:

https://udm14.com

oof.

Microsoft Should Make A Version Of Windows That Doesn’t Suck And Then Gets Worse 

https://aftermath.site/microsoft-windows-11-debloat-copilot-ultralite-good-version

  Quote

Now, I can see you nerds out there typing “Chris, you could simply install Linux,” and you would be correct. But let’s just table that for a second, this is between me and “M$” as we used to call Microsoft back in my day. How about you create a version of Windows that’s just for me, without any of those new features you’re introducing to confuse older people with poor computer literacy? Bing news results and ads every time I try to look for an app? Don’t need it.  An AI Copilot that isn’t particularly helpful and that I find morally repugnant? Pass! OneDrive installed by default? No thanks, I am an adult. And this new hardware-enabled feature that takes a snapshot of everything I do and say at all times? Oh wow, that’s really bad, man, why would you think I would want that on my computer? None for me!

Expand  

I objected and literally just wrote, "Stay out of my stuff" and received this

  On 5/24/2024 at 9:59 AM, Squee said:

I objected and literally just wrote, "Stay out of my stuff" and received this

Expand  

"stay out of my girl"

  On 5/23/2024 at 3:24 AM, usagi said:

she's advertising AI products in this very video, wtf?

yeah, it's not entirely against the concept, more certain business practices / empty promises. hence the 'big tech'. there's quite a big section about open source solutions towards the end.

just thought it might be interesting hearing critique coming from an insider, so to speak.

  On 5/23/2024 at 2:16 PM, Rubin Farr said:

Expand  

lol

  On 5/23/2024 at 8:04 AM, usagi said:

the internet itself, which [...] continues to rewire our monkey brains and alter society in unexpected and not entirely good ways.

watmm

  On 5/24/2024 at 2:13 PM, usagi said:

"stay out of my girl"

  On 5/24/2024 at 7:21 PM, decibal cooper said:

Expand  

This might be funny, but it is also incredibly sad how (sorry for the words) fucking stupid Sillicon Valley is getting, not because AI spitting complete nonsense, but because they are still pushing this useless garbage as the next big thing. I decided to cancel my sub to ChatGPT and will try Claude, but actually I'm starting to avoid anything AI related by now.

Just to add a bit, I feel that desperation to just fucking be the numbah one instead of actually creating USEFUL technology.