Red Letter Media

  • 🐕 I am attempting to get the site runnning as fast as possible. If you are experiencing slow page load times, please report it.

Favorite recurring character? (Select 4)

  • Jack / AIDSMobdy

    Votes: 223 23.6%
  • Josh / the Wizard

    Votes: 65 6.9%
  • Colin (Canadian #1)

    Votes: 414 43.9%
  • Jim (Canadian #2)

    Votes: 200 21.2%
  • Tim

    Votes: 352 37.3%
  • Len Kabasinski

    Votes: 190 20.1%
  • Freddie Williams

    Votes: 244 25.9%
  • Patton Oswalt

    Votes: 21 2.2%
  • Macaulay Culkin

    Votes: 472 50.1%
  • Max Landis

    Votes: 52 5.5%

  • Total voters
    943
Just to add a point I haven't seen people discuss before but I like AI because it allows me to circumvent copyright. You want to use a song as a background in a video you're making but you don't want your video to get a copyright strike? Just ask the AI to make a legally distinct version of that song and here you go. Same thing goes for images, videos etc. AI will be the death the intellectual property rights because it can create new versions of things that already exist.
 
And sure enough
Screenshot 2025-05-28 191532.webp
 
I hate the uncanniness of AI video. As fast as it's developing there's always that inhuman factor that creeps me out. I'm always expecting to see faces and limbs transmorphing into something else like they're made out of water.

I love messing around with AI for writing and arts, but video just doesn't cut it for me, at least not yet.
 
These guys still see themselves as big-time moviemakers who just haven't been discovered yet. They're scared of AI because they're afraid it'll take the jobs they want without realizing that it could enable them to do the jobs that they currently have even better.

The current wave of visual generative AI doesn't do anything now that couldn't be done before -- it just makes it much MUCH easier to do. Over 20 years ago, a team of animators using supercomputers made the uncanny valley Tom Hanks character in The Polar Express movie. Now any doofus with a laptop can do the same in minutes with the help of AI.

This should be seen as a tremendous boon to the middle-aged kids of Milwaukee. How many times in HitB or BotW have they light-heartedly "pitched" a movie idea or said "Here's what they should have done" when talking about a film? The technology now exists for them to create their hypothetical scene/trailer/whatever on the fly and drop it into their videos. It would be a tremendously useful tool for them if applied correctly.

I don't really have an opinion on the bigger picture regarding generative AI, but for the RLM crew specifically, I'm surprised that they're not enthusiastically embracing it.
 
These guys still see themselves as big-time moviemakers who just haven't been discovered yet. They're scared of AI because they're afraid it'll take the jobs they want without realizing that it could enable them to do the jobs that they currently have even better.

The current wave of visual generative AI doesn't do anything now that couldn't be done before -- it just makes it much MUCH easier to do. Over 20 years ago, a team of animators using supercomputers made the uncanny valley Tom Hanks character in The Polar Express movie. Now any doofus with a laptop can do the same in minutes with the help of AI.

This should be seen as a tremendous boon to the middle-aged kids of Milwaukee. How many times in HitB or BotW have they light-heartedly "pitched" a movie idea or said "Here's what they should have done" when talking about a film? The technology now exists for them to create their hypothetical scene/trailer/whatever on the fly and drop it into their videos. It would be a tremendously useful tool for them if applied correctly.

I don't really have an opinion on the bigger picture regarding generative AI, but for the RLM crew specifically, I'm surprised that they're not enthusiastically embracing it.
"AI, please generate a 90 minute film sequel to Space Cop, but make Mike Stoklasa's character fatter and dumber, and make Rich Evans's character resemble the gigachad meme."
 
These guys still see themselves as big-time moviemakers who just haven't been discovered yet. They're scared of AI because they're afraid it'll take the jobs they want without realizing that it could enable them to do the jobs that they currently have even better.

The current wave of visual generative AI doesn't do anything now that couldn't be done before -- it just makes it much MUCH easier to do. Over 20 years ago, a team of animators using supercomputers made the uncanny valley Tom Hanks character in The Polar Express movie. Now any doofus with a laptop can do the same in minutes with the help of AI.

This should be seen as a tremendous boon to the middle-aged kids of Milwaukee. How many times in HitB or BotW have they light-heartedly "pitched" a movie idea or said "Here's what they should have done" when talking about a film? The technology now exists for them to create their hypothetical scene/trailer/whatever on the fly and drop it into their videos. It would be a tremendously useful tool for them if applied correctly.

I don't really have an opinion on the bigger picture regarding generative AI, but for the RLM crew specifically, I'm surprised that they're not enthusiastically embracing it.
There’s truth here, but if they even joke about doing this, their audience will eat them alive.

Then people who’ve never heard of them, but suddenly know everything about them on Twitter and Reddit, will join in. It’s absolutely verboten in that space to even do that stuff for private fun, let alone turn it into content.

They’d get less flack punching an old woman at a store.
 
I hate the uncanniness of AI video. As fast as it's developing there's always that inhuman factor that creeps me out. I'm always expecting to see faces and limbs transmorphing into something else like they're made out of water.

I love messing around with AI for writing and arts, but video just doesn't cut it for me, at least not yet.
It takes me longer to get that feeling, but even now, there is a hyper-fluidity to motion. I guess that's the next step, to create believable hesitation and slight awkwardness.
 
There’s truth here, but if they even joke about doing this, their audience will eat them alive.

Then people who’ve never heard of them, but suddenly know everything about them on Twitter and Reddit, will join in. It’s absolutely verboten in that space to even do that stuff for private fun, let alone turn it into content.

They’d get less flack punching an old woman at a store.

You're probably right. People have passionate feelings about AI content, and I could envision a big backlash even if they used it in a responsible way that compliments their original commentary. Maybe attitudes will change as the technology becomes more common, but who knows... there are still members of the photography community who use "Photoshop" as a pejorative term (even though all pictures have been touched up in some way since the dawn of photography).

Who are these fake trailers even for? Do normies find this entertaining? It just looks uncanny and shitty.

I wondered the same thing as I was watching their video. Nearest I can tell, the trailers are just clickbait slop. Titles like Back to the Future 4 and Titanic 2 pique the viewer's interest when they see the thumbnail in their recommendations, and they get a click. Grandmas think it's real and share it on Facebook, and it gets more clicks. Those clicks add up, the channel is monetized, and the channel Mike mentioned pinches off content every day. That sort of volume adds up, and somebody's making money off of it.

BUT, it won't last. Remember that one of the key points is that it's all very low effort. When you get anything super easy to do, a whole bunch of people are going to try to do it. The fake trailer space will get saturated very quickly as more grifters try to get in on the scam. Soon these fake trailers will be as easy to ignore as "Hot Singles in Your Area" ads.
 
Even with their next-to-zero knowledge about AI (same as us all, probably) there are a lot of interesting things to think and talk about. Mike in particular: shouldn't these sort of ethical dilemmas and human considerations be right up his alley?
but why do they have zero knowledge? they've been making over a million from patreon for over a decade now, what the fuck are they doing with their time? its the same shit with movies, why do they only have the time to attend movies on thursday night at the blackest theater in town?

even if you don't know anything, its not that hard to research, spend an hour or two to get caught up. especially in the last decade where everyone is obsessed with making short form content it shouldn't be difficult to learn about anything.
a human can (is supposed to) conceptualize and action visual concepts that are not explicitly shown onscreen
"is supposed to" is right, and that's the thing, when you look at blacks or indians (a majority of the human population) they are less capable than even AI when it comes to visual concepts and conceptualizing and yet they obviously make films in india and nigeria so would you consider them in the same field as AI made entertainment? the same level of goyslop?

and what if you gave a script of something an AI made and had humans make that and you directed it and released it, is it still AI written based on how directors naturally switch up stuff during shooting that may not have been in the script? what about the adlibs?

And whats stopping a human from being an intermediary entirely, the same way certain white people would buy a house and be the "face" the bank sees when in reality it was a black family back in the days of redlining. there were plenty of banks that assumed they were interacting with a fellow human instead of a human doing the evil misdeeds of cotton picking farm equipment
Real art has passion in it that a sensitive person can see. It may not be flawless or even very good, but you can see the love it was made with and it transforms it into something more. It's the difference between a million mass produced Barbies and the little, hand-stiched ragdoll made by a great-grandmother for her first great-grandbaby even though her hand shakes and see can't quite see like she used to.
except what if there is no passion to it, the "in it for the money" acting job, there's a reason you're seeing tons of AI video prompts of people making fake drug commercials or news broadcasts, AI is amazing at making goyslop, like those Man on the Street interviews.

the second its more involved than the same goyslop you're having shoved in your face it fails completely,
Either way, there had to be a person behind the scenes coming up with the prompts, so is it not still an act of creativity?
great point, another point to add is that people have always said the same shit about new tech, moving from Cels to digital animation or 2d to 3d always leads to people trying to say it "doesn't count" or "isn't creative". they said the same shit about rap music and cover songs because it "wasn't original" and while people might still have that view, enough people didn't give a shit and the newer generations grew up with the tech to the point where they'd think you gone mad if you demanded cel shading and animating on ones instead of how they typically do shit now.

Especially when so many people claim George Lucas' wife is the true creator of star wars because she edited it, isn't someone prompting AI just basically editing from an unlimited reel of footage? If they're going to say all the women that edited movies are their real creators, then why can't you say the same thing about the lady using chat GPT instead of an editing bay.
The thing to fear is AI creating fake videos intended to influence elections or start wars.
remember when Benghazi happened and they blamed it on a youtube video? according to the democrats you're about a dozen years late to fearing AI created videos causing wars.
If only Hollywood had a tenth of such passion or spirit.
they did, hell they still do. its just that you never see it in big budget pictures,
Considering Spielberg became a pablum-peddling hack once he hit it big, calling a flash in the pan like Abrams the next Spielberg is pretty accurate.
say what you will about Spielberg he made sure to pay it forward, so much of beloved 90s kids stuff is because of him, same thing with George Lucas, they got in power and tried helping the next generation of creators. Same thing with guys like Tom Cruise or Adam Sandler or Rodney Dangerfield or Quincy Jones or Dr.Dre or Lloyd Kaufman. You take what you have and try and help everyone else. Meanwhile how many people can attribute their success as a behind the scenes creative from JJ Abrams giving them a shot? Where's JJ Abrams' Animanics or Monkey Island? it seemed like every fucking comedy in the late 2000s was made by Judd Apatow because once he got successful he helped greenlit all his buddies ideas.
? Just ask the AI to make a legally distinct version of that song and here you go
can it do that?
I hate the uncanniness of AI video. As fast as it's developing there's always that inhuman factor that creeps me out. I'm always expecting to see faces and limbs transmorphing into something else like they're made out of water.

I love messing around with AI for writing and arts, but video just doesn't cut it for me, at least not yet.
thats because of the Uncanny Valley effect, it will never truely be able to overcome that if you ask it to make humans, but the obvious use case is in animation, literally the shit people are making but involving talking cute animals or space aliens demolishes the highest budget dreamworks or disney shit. by the time shrek 5 comes out you'll be able to make a better shrek in your own living room.

it reminds me a lot of Jak&Daxter and Ratchet&Clank, both studios admit they did aliens instead of humans like other studios not only for the kid aspect but because humans are way easier for people to judge and critique while no one can go "that's not how otsels walk"
Who are these fake trailers even for? Do normies find this entertaining? It just looks uncanny and shitty.
its a joke, this movie had a legal issue but was made entirely before the current AI revolution
It takes me longer to get that feeling, but even now, there is a hyper-fluidity to motion. I guess that's the next step, to create believable hesitation and slight awkwardness.
what i don't like is how it still can't get fluids right, or how objects interact, a baseball hitting a piece of cardboard. i don't think it can ever get motion down too well. like those AI videos of sitcoms or comedy shows, it can't do the subtle movements, such as two people shaking hands or doing anything too intricate. Another big point is that everyone looks so movie quality. i've seen multiple videos of people parodying AI and its super easy to tell because even if they're attractive, they still have the marks and bumps that humans have.
 
The biggest issue with AI I see right now is movement.
Unless it's either a pre-choreographed sequence or really slow movement, AI will not be able to do it.

The truth we need to acknowledge is that AI will never know how a real living being looks like or moves.
When you tell AI "create a human" but the internet is filled with cartoon humans and fictional characters who have superpowers, AI will not be able to differentiate between real and fiction.
It will treat Spider-Man's movements as real human movement and will incorporate his abilities into the creation.

If I spam 100 billion pics of humans with 2 heads online, that will change the way AI draws humans.
That's because AI is not smart, it's a blender and if you blend feces and piss along with everything else, all creations with look and smell like piss and shit.

This is what will ultimately kill AI years in the future.
The perpetual degradation of all AI creations until you can't make anything good anymore because 95% of things that AI draws from are already AI creations based on AI creations based on AI creations.

I'm not even sure if we can reverse it, not on the current internet.
I've said this before but AI will become so prevalent online within the next 10 years and outnumber real humans by such a huge amount that we will have to abandon it and create a new internet.
 
The biggest issue with AI I see right now is movement.
Unless it's either a pre-choreographed sequence or really slow movement, AI will not be able to do it.

The truth we need to acknowledge is that AI will never know how a real living being looks like or moves.
When you tell AI "create a human" but the internet is filled with cartoon humans and fictional characters who have superpowers, AI will not be able to differentiate between real and fiction.
It will treat Spider-Man's movements as real human movement and will incorporate his abilities into the creation.

If I spam 100 billion pics of humans with 2 heads online, that will change the way AI draws humans.
That's because AI is not smart, it's a blender and if you blend feces and piss along with everything else, all creations with look and smell like piss and shit.

This is what will ultimately kill AI years in the future.
The perpetual degradation of all AI creations until you can't make anything good anymore because 95% of things that AI draws from are already AI creations based on AI creations based on AI creations.

I'm not even sure if we can reverse it, not on the current internet.
I've said this before but AI will becomes so prevalent online within the next 10 years and outnumber real humans by such a huge amount that we will have to abandon it and create a new internet.
That is all because machine learning models are NOT AI, just bulk data processing.

The main thread I see in people afraid of AI is that they think things like ChatGPT are actual AI, rather than just data aggregators with advanced filtering.
 
can it do that?
Well, yes and no. (i.e., the question was whether you can generate legally distinct AI songs / art works to avoid copyright issues in relation to the original)

In practice yes because you can already just generate something "in the style of X" and often time this will be enough to not get an automatic copyright strike from a website like YouTube because the algorithm won't be able to recognize it as the protected work (provided you changed it enough). The moment you start monetising your work though, it may be a different story.

Legally, it's a tricky question and it's up to the courts to determine what exactly is "legally distinct AI". If you regenerate a scene from a movie entirely, keeping vaguely the same script, I think it would be possible to avoid copyright issues. With music, you need a license to play a cover of a song (i.e., a mechanical licence), but you are generally free (with certain limits) to play a different song inspired by the first song.

The question is wehether something is a rip-off, a derivative work or different enough to qualify as a new work. Sometimes, a derivative work might cause copyright infringement of the original (but not always). For example, you can create a remake of Romeo and Juliet because it's in the public domain but you're not free to create a remake of Star Wars which is not in the public domain. On the other hand, you are free to create a new work inspired by an existing work. This would be the case of the movie Spaceballs from 1987 (noting however, that George never sued Mel Brooks for this one because the two had an agreement and also the fact that parodies have special protections against lawsuits in the US). The question is generally whether there is enough authorship and originality in the new work to make it "transformative". Even a derivative work can be transformative if it alters the original work to such an extent that it creates a new work with its own meaning or message, potentially avoiding copyright infringement.

Some people have been trying to argue that if you feed the AI something which is protected by copyright, then this is already copyright infringment. I personally think that this is impractical and goes against the developement of AI technology. They are basically saying that you cannot train your AI model on things that are already protected by intellectual property rights. In practice, this is not established and under "fair use", you can use an existing protected work if the output you create is sufficiently transformative as to not replicate the original (which ai usually is).

Tldr: Yes, you can generate AI songs that avoid copyright issues if they are substantially different from the originals (i.e., "in the syle of"). AI-generated music "in the style of" an artist may pass algorithmic detection and avoid strikes; however, legally speaking, copyright infringement depends on whether the new work is considered transformative and original enough to not qualify as a rip-off / copy etc. Parody and inspiration are often protected, but using copyrighted material directly or too closely may require licenses. The legality of training AI on copyrighted works is not settled. Some people argue this is covered by fair use if the output is original enough.

EDIT: I'm not an attorney and this is not legal advice

the truth we need to acknowledge is that AI will never know how a real living being looks like or moves.
When you tell AI "create a human" but the internet is filled with cartoon humans and fictional characters who have superpowers, AI will not be able to differentiate between real and fiction.
It will treat Spider-Man's movements as real human movement and will incorporate his abilities into the creation.
I don't agree at all. This all depends on how you train your model and in fact, that is already what distinguishes bad models from good models. AI models are designed by people who select what goes into the model for training purposes. You don't just throw in everything that exists on the Internet. The result would be complete nonsense
 
Last edited:
I don't agree at all. This all depends on how you train your model and in fact, that is already what distinguishes bad models from good models. AI models are designed by people who select what goes into the model for training purposes. You don't just throw in everything that exists on the Internet. That would be complete non-sense
"Nonsense" shouldn't have a hyphen in it.
 
These guys still see themselves as big-time moviemakers who just haven't been discovered yet. They're scared of AI because they're afraid it'll take the jobs they want without realizing that it could enable them to do the jobs that they currently have even better.

The current wave of visual generative AI doesn't do anything now that couldn't be done before -- it just makes it much MUCH easier to do. Over 20 years ago, a team of animators using supercomputers made the uncanny valley Tom Hanks character in The Polar Express movie. Now any doofus with a laptop can do the same in minutes with the help of AI.

This should be seen as a tremendous boon to the middle-aged kids of Milwaukee. How many times in HitB or BotW have they light-heartedly "pitched" a movie idea or said "Here's what they should have done" when talking about a film? The technology now exists for them to create their hypothetical scene/trailer/whatever on the fly and drop it into their videos. It would be a tremendously useful tool for them if applied correctly.

I don't really have an opinion on the bigger picture regarding generative AI, but for the RLM crew specifically, I'm surprised that they're not enthusiastically embracing it.
I don't think it's that hard to figure out — the people most excited about it are the people with zero developed skills, because they can't make anything without it and they don't see all the shit that sucks about it. The more you know about any particular art form the less AI slop appeals to you. The ignorant are dazzled more easily.
 
If I spam 100 billion pics of humans with 2 heads online, that will change the way AI draws humans.
That's because AI is not smart, it's a blender and if you blend feces and piss along with everything else, all creations with look and smell like piss and shit.
The funniest thing about this statement is that AI pictures are becoming noticeably more and more yellow.
 
Back