The Future Direction of Nightmare Masterclass

[@jeniferjoseph9200]

I want this to come from a place of knowing that I basically agree with you in everything. I found you from your horror/strange media content and I prefer that content to free-form discussion of social topics. A lot of comments like this new direction and if you’re finding success with that format that’s great. What I will say is there can, if you want, be both.

[@tserov2336]

Discovered this channel through the Petscop investigations, sure, but as a mathematician, I've always been outside the humanities (and to a large extent philosophy) looking in. What I found uniquely appealing (not just) in the Petscop Investigation was the comprehensive application of - from my perspective - the "humanities"; other theories and viewpoints I'd never heard of... ... you blew my mind, dude. And I'm here for that...

[@mushmushmush]

I'm really glad for these new types of videos! I loved your analyses but hearing your perspective in this more ramble-y format is super interesting! You are incredible at putting complex ideas into words!

[@SendyTheEndless]

I don't dislike AI because the tech bros made it, I dislike it because WE made it, and it's sold back to us as proprietarily owned "magic". Because it's here to replace human thought processes and colonize the mindscape. It's also being pushed on me in every form of tech and it's made me pine for the old days of personal computers. NO! I don't want an AI "buddy" to "help" and spy on me. Still subscribed though. I'm glad to hear your thoughts on it.

[@PuppyLuver256]

Said a lot better and more concisely than I did, you get a like c:

[@NightmareMasterclass]

This sounds like the language of a moral panic. I'm not on board.

[@SendyTheEndless]

@@NightmareMasterclass Well, yes, it is something of a panic. Let's just say I hope your cautious optimism convinces me : )

[@NightmareMasterclass]

@SendyTheEndless Hoping to articulate myself a little better on the topic in some upcoming videos.

[@randomjunkohyeah1]

@@NightmareMasterclass I don’t know what to tell you- some things just _are_ worthy of panic. At least a bit.

[@samuraiswordbrub4642]

i found ur channel thru the petscop videos back when petscop was still coming out but i found that i really liked your analysis on things in general and ur videos have been consistently good since then. i cant wait for the future of this channel. thanks for all the work david!!!

[@RyderC]

feel the same way

[@MrXwingzerox]

Keep doing what you’re doing. I appreciate your content and desire for free-thinking amongst your audience. Cheers bud

[@DeuceRumbo]

Keep the content coming, your analysis is always appreciated. It can feel a bit maddening to have this deluge of media content and to have so few voices that care to even begin to identify capitalism as a core problem in society.

[@DesmondDuval]

I like The Villain Arc, you've already got the goatee... I came here for the ARG and surreal horror stuff, but 11/10 for living your truth and telling the haters to get fucked. Rock on David!

[@SendyTheEndless]

It's almost asif, these people are cool with you talking about these online horror and surrealist projects, but the moment you begin to notice how patterns in them mimic or rhyme with patterns in the real world, there's a sense of unease and almost violence associated with that kind fiction coming outside of the computer : )

[@elmalifico3708]

2:14 I don’t know if you’re familiar with Duncan Trussell, but I used to listen to his podcast and I think this was something he tried to communicate as well. His goal was to try to create novelty, but you may have a different goal. But I’m here for the ride and see where it takes us.

[@iridelombardi]

im skeptical of atleast this form of ai but its good to have the vibes based assumptions counterargued :3 about consciousness, do you know about attention scheme theory? that is an interesting line of research on consciousness! thanks for your vids!!

[@iridelombardi]

the author of that theory argues that thinking that consciousness is special and emergent from complexity and not a specific mechanism that science could investigate is still a sort of magical thinking

[@NightmareMasterclass]

@@iridelombardi I'll do you one better and say that "consciousness" in the way that most people talk about it reveals a sort of magical thinking that attempts to distinguish humans from everything else.

[@iridelombardi]

@@NightmareMasterclass the way we perceive consciousness as magical might be a priori knowledge ,yes we can think rationally that is is still computed by the brain but we can’t really perceive it as non magical : tha hard problem of consciousness, qualia, etc

[@NightmareMasterclass]

@iridelombardi I'll spend some time criticizing the "hard problem of consciousness" in my upcoming analysis of 3D Workers Island. I'll have the section timestamped.

[@Grogeous_Maximus]

Good to see you cheerful, after all of the politican doom and gloom of 2024. Personally, I'm a grinch, but happy New Year, mate

[@johnminser4935]

Since there’s a lot of AI discussion in the comments and I think at one point you said your analytical lenses include Orthodox Marxism, I’m wondering how you feel about the long term economic impact of AI. Essentially, it democratizes the ability to produce competent-looking art but doesn’t teach the techniques to create that art. Over time, this produces incentives to create AI art, and so the incentives to learn art decline. Over 5-10 years, people will produce massive amounts of AI art but comparatively less human art, leading to AIs training on AI-produced art. At worst this leads to instability in the production system and at best to deeply homogenized art that can only train recursively. Do you think that there’s a risk that, with a smaller pool of trained artists and therefore an even smaller pool of people with an interest in experimentation, that art as a medium incorporating new styles shrinks and art as a homogenized commodity further becomes the norm?

[@NightmareMasterclass]

I have never been concerned about the number of so-called "trained artists." Lol

[@johnminser4935]

@@NightmareMasterclass powerful implications lol - would love to unpack but don’t want to take up your time and anyway YT comments not the place for it. Cheers and thanks for responding on your holiday!

[@NightmareMasterclass]

@johnminser4935 You too! To be clear, this does merit a longer response but I'm short on time at the moment. Perhaps a future video topic for 2025.

[@PuppyLuver256]

I'm gonna be honest, most people who are criticizing genai programs aren't doing so because "the tech bros made it". Hell, a lot of us aren't even criticizing it for the very valid environmental reasons, or at least that's not the main reason. I and other artists are criticizing it because _we don't want our work that we've poured countless collective hours into scraped for some computer to put into a metaphorical blender and shit out a bland plastic-glazed turd without our consent_ . It's bad enough when other human users are stealing and reuploading our work under their name or by hand-waving with "credit to the artist uwu " or "idk who made this I found it on google" or "dm for credit / removal", the scraping is just making it that much worse for us as it's harder to tell what all went into the blender than it is to locate and take down unauthorized reuploads of single unaltered or barely altered pieces. The point you're trying to make about consciousness or lack thereof is completely irrelevant, we just don't want randos takin' our shit that we worked our asses off to create. My feelings would be the same whether or not genai was proven to be "conscious". After all, the random dude who reuploaded my art into a yt short _and kept doing so even after repeated dmca claims on my part_ last year was definitely conscious, if a moronic asshole. That being said, good on ya for the trans allyship. We may have disagreements on the appropriateness of genai and different understandings as to the core reasons why people tend to be against it, but we're deffo on the same page when it comes to marginalized people being allowed to friggin' Exist and be happy. (not to mention as a nonbinary person and thus under the trans umbrella, any attack on trans people that isn't specifically a case of transmisogyny is also an attack on me, and that's if they don't just flat out deny enbies are real in the first place)

[@NightmareMasterclass]

In general, I would advocate for public ownership of these systems with artists always having the option to opt out if they so choose. Pragmatically speaking, here's the truth. At best, maybe you'll get some kind of legislation that allows you to opt out or collect some nominal royalty amount. After all, the degree to which any one work contributes to this training process is miniscule.

[@PuppyLuver256]

@@NightmareMasterclass Unfortunately I've seen firsthand that merely giving artists an option to opt _out_ isn't a good solution either, as more often than not by the time that artists are given that option their stuff has already been fed into the training data. Sure, opting out will signal to the system to stop collecting from those artists, but what about what it already grabbed before those artists knew about it? From what I understand it can't just be taken out without deleting the whole set, if even _that's_ entirely possible. And what's stopping current owners of genai systems from simply not honoring an opt-out request? Or worse, removing the option entirely because the ceo's a whiny manchild who has gone on record with his hatred of the very idea of having to credit artists for their work and is upset that most people won't actively let their work be chewed up and spat out by his oh-so-creatively-named snark machine that he constantly had to lean on to keep from """going woke""" (at least until a lot of the so-called ""woke"" left the site he trashed)? I will concede that having these systems be publicly owned could potentially alleviate some of the above issues I have with how genai systems are currently operated, as would running on an opt _in_ approach to training data collection rather than the current grab-first-ask-permission-later opt out method (which of course has its own problems if the person who opted in is a reuploader and not the actual artist, but lbr art thieves are gonna be a problem for as long as art exists). I'm sure that some people behind these things are even trying their best to work on making their operations more environmentally friendly as well, although I'm not exactly holding my breath on that. Of course even if all of those issues were solved in the best way possible for all parties I still wouldn't be too fond of the results, if the phrase "bland plastic-glazed turd" I used earlier was any indication, but that's the _least_ of my concerns and a matter of personal taste. (and tbh even the minuscule contribution of any one work still skeeves me out if that work wasn't added with full knowing permission from its creator, kinda like how if I know there's a non-zero number of spiders in my house at any one point it's not exactly reassuring to tell me that any spider that may be in here is actively trying to remain hidden from view)

[@NightmareMasterclass]

Honestly, this fixation on copyright and regulation is not very interesting to me, and I don't think it will get you anywhere either. I've been around long enough to know that any legislation that is passed will inevitably be pro-corporate.

[@ekiouja]

A lot of the discourse around ai is a result of said tech bros selling it as a tool for everything. It has been shoehorned into too many applications where it where it is generally less efficient than existing systems. It is has its use cases however the current amount of industry hype is primarily there to take advantage of venture capital firms. Ai is a tool, and it is currently being used by the capital class to justify further corner cutting at the expense of workers and consumers. That isnt ai as a tools fault.

[@NightmareMasterclass]

Yep. Well said.

[@wikikomoto]

oh... wow okay, i wonder if you've actually even looked into AI at all?? like, i'm hoping you're coming from a perspective of ignorance and you're not intentionally avoiding all the primary reasons people "don't like it." the reason people are opposed to AI mostly consists of it taking work away from professional artist, and that its based almost entirely off stolen art used without permission. to completely assign its dislike to something to do with "tech bros" and comparing its environmental impact to vehicles (something people need to get around, regardless of work, where there is literally zero reason for AI, and we existed fine as a society without it 5 years ago) is just... such a weirdly defensive position to have. and to dismiss its human impact in the moment, right now, is quite crazy too. from any type of a progressive standpoint, there is zero benefit to AI being used to replace skilled artists. I'm a professional artist myself, and most of my friends are. and i was very skeptical of it taking work away from us. i thought that perspective, even just 2 years ago, was unrealistic and alarmist. yet, here we are in 2024, and i haven't talked too a single artist who hasn't been negatively effected by AI. i know four people personally who have lost their jobs to it. two traditional artists a writer, and a photographer. someone who never even knew their job was even AT RISK of being replaced by AI. how could it??? yet here we are. already. barely a half decade since its introduction. a research firm i used to work with shutdown because of it, and they were one of the few research groups i knew about that released their information publicly to academics without a fee. and these are just my personal experiences of AI having tangible, real world negative impacts. right now. and not some nebulous threat in the future. in a theoretically better society, AI would be used to replace menial tasks. to free up more opportunities for people to pursue art. and yet, its being used for the EXACT opposite. replacing artists, to force them to find non-artistic means of income. i just don't get how you can ignore all that, because you think its "kinda neat" like, yeah it is kinda neat. and you can enjoy it, and even use it. most people do now. thats fine. but to completely ignore ALL its negative real-world impacts its having RIGHT now, and assign its "dislike" to leftists thinking "tech-bros" are lame is just... its more than reductive. its either willfully ignorant or intentionally malicious

[@NightmareMasterclass]

@wikikomoto I'm not expressing myself clearly enough in this video. This was not scripted, and the way I framed it was perhaps a bit too defensive. But you have a false impression about my views. To clarify, I do not generally approve of the way AI's being used currently. But that's not the technology's fault. As you have said, if it was employed to benefit society, it could actually make everyone's lives a lot easier. I support public ownership of the technology. It is trained on the public's data. It should benefit the public, and artists should be able to voluntarily choose if they want models to be trained on their work. Here's the reason for my defensiveness. There is a double bind here for me. On one hand, if I talk about just how advanced the technology has gotten, I will be maligned by AI skeptics for "buying into the hype." Even though, as you said, there isn't a person in your field who hasn't been affected by this. I am certainly not suggesting it is generally as good as a professional artist at this point in time. But it's rapidly improving in many domains, and it's already displacing jobs to some degree. On the other hand, if I ignore or downplay these advancements in an effort to placate AI skeptics, that would be insincere. I've come to believe AI's development has much broader societal implications. And all I really want is the space to think openly and honestly about it. I will state my views right off the bat in the future to prevent this kind of confusion.

[@wikikomoto]

@@NightmareMasterclass wow i replied yesterday with a really long comment but i guess youtube decided to just not post it. oh well. mostly I just wanted to say thank you for the well thought-out reply. and i hope to hear more of this perspective in the future. i don't think AI is inherently evil either. i personally haven't heard any of the "discourse" around tech-bros and the like. i think AI has the potential to save lots of lives with its use in medical fields, and problem solving through its use in mechanical design. its just a shame to see it almost entirely being used in its current form to replace human workers with cheaper, easier to manage alternatives. and again, in fields nobody suspected it would threaten (photography for instance) i just don't personally have much hope for any big company stopping its use anytime soon, and seeing it replace artists here on youtube for basic stuff like thumbnails and narrations has been.... just, not fun to see.

[@NightmareMasterclass]

@wikikomoto I'm glad you took the time to hear me out. Sorry, I'm not sure why your comment was not posted. I'm not seeing it as held back on my end. KZbin is really annoying sometimes.

[@wikikomoto]

@@NightmareMasterclass i'm glad you took the time to write out a reply that directly addressed my issues. thank you~

[@zab416]

I'll hear you out, but something something universe turning into paperclips. I guess I fear it and am annoyed by AI for some reasons that I feel are are valid and some that I know are pretty patently ridiculous. Not so much skepticism in my side eye.

[@1dr_wall769]

As a trans person with a degree in data science/machine learning, this ramble was laser focused at me 😂 I am an AI skeptic, but that comes from seeing the capabilities of the technology corrupted and coopted by the tech bros that can only think of it as another form of capitalistic exploitation. I think that AI and machine learning really does have the capacity to democratize a lot of information and learning that was previously only accessible to the elite, in the same way that the printing press and the internet did before AI. The problem comes from the flawed training of these models: all the problems that people have with AI (that I've heard, at least) are actually problems with how models are trained, not with the tech itself. Like, the models are trained using art stolen from a​ctual humans, or the models are only trained with data from people the tech bros think are relevant to their bottom line (read: straight white cis men). If the access of and education about this technology is democratized, I believe marginalized people could use AI to level the playing field and increase the general education of the masses, and ideally it could be the first domino in a chain to liberation, like the printing press was for women's suffrage

[@NightmareMasterclass]

The "stochastic parrot" argument is by far the most popular high minded critique of LLMs, and it is fundamentally about the technology itself, not the way it's trained. So, I'm surprised that's not on your radar.