@@tadasgame I'm just 30 years old and do I really have to say it? back in my days this would be considered shit posting, I've just watched 22 minutes of numb dumbing content. "make me fat "haha here is a stock photo of a fat kid with weird edges that even an newb photoshoper wouldn't add "epic! XD" I had to stop it at 22:37 this is seriously what makes you laugh? "The grape lady" was funnier then this whole video? I'm pretty sure that even Jack would start laughing just by hearing the sound of that ancient video. I could sample 5,10,20, 30 minutes of this video and nobody would laugh at just the audio. "I haven't laughed this much in a while" that's just depressing
@@svampebob007 I'm 34 years old and I'm sure I have the same exposure to the internet history as you, if not maybe a little bit more for having been around 4 years longer. Bro, what the hell are you saying? Let people enjoy what they enjoy, that's what the internet is about. Shitposting doesn't exist in objectivity. Most of GenZ content is "shitposting" to me, because I don't get it/enjoy it. Most of contemporaneous art is also shit for me, because I don't get it. But the universe doesn't revolve around me and my preferences, so I know not to judge what other people enjoy based on my own preferences. Unnecessary, man, totally unnecessary. Keep the judgement to yourself.
The "teeth" prompt generating toothbrushes is kinda interesting, because it reveals something interesting about he dataset the AI was trained on. Within the Data set, probably most of the pictures with people in them also had teeth in them, but it's not something you, as a person, would think to tag and point out to the AI. However, on pictures that have to do with oral hygiene or dentistry, you would feel inclined to add "Teeth" to the pictures tags. In the training process the AI then looks at those data sets and concludes that the difference between pictures with and without "teeth" is stuff like toothbrushes and wide open mouths. The data set had a human perception bias, almost like a blind spot, and as a result the AI doesn't know what "teeth" are.
@@hepatitisf7495 mind you i don't train AI or anything even resembling it, but i'd imagine you'd have to go through and tag all images with teeth in them as teeth, otherwise the AI would start assuming that open mouths or other parts of the mouth are teeth. i'm guessing it needs some of that outside consistency otherwise it'll confuse itself
they could pull a feather out of a certain sodium websites hat and imply tags based off of more general things and maybe also add a system to remove certain implications if they arent in the image
The photoshop image diffusion model is quite terrible compared to what's out there (and I image the problem of overfitting that you mention could be solved with, in order of descending difficulty: 1. a superior image-to-text model during the training phase; 2. Adopting "Low Rank Adaptation" techniques to fine-tune the model for specific problem areas. This method would avoid the need for retraining the entire neural network from the beginning; or 3. even just by using negative prompting to further refine the output). It is true though that what you're referring to is widely acknowledged as a big problem with the 'big data' strategy modern AI research and development employs. However, it's been said that the future of AI may well lie wuth the 'small data' approach, i.e. quality over quantity. With this approach researchers might, for example, use small, highly curated datasets to finetune a larger model such that it's broad knowledge can be transferred to something precise and accurate, eventually creating a conceptually perfect text-to-image model in time. Although we first need to learn how to crawl before we can walk.
It's almost as if GIGO applies to data sets and if the data inputs that it learns with are garbage and not carefully vetted, then guess what the results are going to be??? Haha
@stephaniegalliart859 that's a harsh conclusion based on ONE case, but it could be true although "angry hispanic" isnt really a stereotype ive ever seen
Some explanations of how Photoshop Generative Fill works that he must've missed: 1. The image completely re-generates whatever is within the selection. Imagine it as the AI "deleting" whatever is in your selection, then re-generating it on it's own. This is why when he selected himself, it completely remade him as a different person. Or why when he selected the whole screen, it generated a whole new image. 2. Correct me if I'm wrong, but I believe that the "prompt" you type before generating an image is what you want the image to be of, not a command that it follows. Instead of writing "make him smile", write "smiling mouth", because you want it to generate a smiling mouth inside the selected area. 3. It doesn't understand logic behind the words written in the prompt, it only sees the keywords themselves. This is why when you for example wrote "make my arms into spaghetti", all that it sees is the word "arms" and "spaghetti", making it generate arms holding spaghetti rather than the arms themselves being spaghetti. And when you wrote "no shirt", it just sees two separate words "no" and "shirt", and since it doesn't know what an image of "no" looks like, it just generates a shirt.
Thank you as someone who understands this stuff I was going crazy on the way he says stuff like "/add/ chicken" 😂 To be fair he's using it like you would chat gpt so it's a fair mistake.
But didn't it understand "remove can"? Maybe this AI was trained with prompts that can represent actions, aside from objects, so to say. It might not understand "convert my arms into spaghetti", but it could get simple prompts more typical of photoshopping, like "remove thing" or "make thing blue". But IDK I haven't watched the whole thing yet, just some speculation, or my thoughts on how this tool would be more useful.
@@jonpatchmodular I wish it did work like that but sadly it doesn't. Like I mentioned, the AI has no idea what is inside your selection (point 1), so if you select a can and write nothing, it's still going to remove the can because it has no idea there used to be a can there. For the same reason "Make thing blue" doesn't work, because it can only generate inside the selection, and doesn't know what used to exist in the selection. For this reason, "blue can" would be a better prompt than "make this can become blue"(although that still probably works because it has the words "blue" and "can" in it) EDIT: Photoshopped has a built in feature where if you write something like "no can" then it automatically makes "can" a negative prompt, meaning that it'll try and avoid it. If your selection is completely covering the can, this is still uneccessary.
" is way funnier than it has any right to be" Does it though? "make me fat" AI: crudely Photoshop fat kid in that looks nothing like him at all, just some AI fat kid and then 22:37 (pause it right there is that funny?) Forced comedy will never beat just comedy gold: "dutch reporter falls off boat into water"
@@svampebob007 okay, we get it, grandpa. youre offended. comedy was better "back in your day" or whatever. is that what you wanted to hear? no one forced you to watch this. if you dont like it, just leave a dislike or something and leave. no need to leave multiple comments berating people for their sense of humor and acting like some kind of comedy elitist freak lmao 💀
I absolutely love the whole AI aesthetic, how it doesn't really understand implied lines, legs, fingers, etc. We see a lot of really creepy undertones in the image where you had the AI fill in the rest of the room. It's like a room, but everything is subtly wrong. There are what appears to be door handles on solid walls, or "box shaped" items that the AI knows we all have, but is severely confused by. Looking at AI images is so reminiscent of ketamine or other disassociatives in a way. Things are a bit alien.
Deepdream turned some of my pictures into something that seems like it can simulate psychedelics. The morphing... the organic look that edges can get. The swirls... the spirals. If you aren't familiar with DD and you love this type of trippy stuff you should definitely check it out.
With the "HAPPY" prompt I think the AI thought that since in many photos of happy people there is confetti and other stuff in the air (covering the face) that it should add that to your face to make it happy
For future, I think it’s better if you understand what the AI is actually doing with your prompt. It is a replacement tool, not an edit tool, it does not get given anything inside your selection and the text promot is a description not an instruction, so if what’s outside your selection does not include necessary context (like skin tone, shirt color, beard) it will choose for itself. This also explains why stuff like “remove X” confuses it, because if you select X then it doesn’t even realise X ever even existed so you have to decide if X would be expected given the context; if it would be, you tell it to not include that (such as “no mouth”, since you’d expect a mouth to be there), otherwise just don’t give a text prompt then it will remove anything “unexpected”.
But... the very first one he did was "remove can." And it did. It removed the can and replaced it with what the shelf (or whatever was on the shelf) might have looked like underneath.
@@KelpTheGreat It would have also done that if he typed nothing. Can’t remember where, but there was definitely other times in the video where that *didn’t* work and he got frustrated, so you’re just cherry-picking
Tears are literally streaming down my face and my sides are in pain from laughter. I downloaded the beta and am going to do this with my friends tonight.
the orange cheetos dust one was my favorite because jack looks like a mildly disgruntled office worker having cheetos dust flung at him by someone off screen
0:05 right off the bat i can already give a tip. photoshop's new ai doesnt accurately produce images from instruction such as "remove this object" its much better to describe what is in the background so the ai knows what you want it to replace the pixels with. "laptop with stickers" or something like that is probably ideal for that scenario
@@Zeffer32 well yes but im just pointing out it’s not the intended way of using the photoshop ai feature. Also there’s lots of points in the video where the ai makes something entirely different to what he expects because he used descriptive language in the prompt instead of telling the ai exactly what he wants to be generated
hi jack, i've been watching your content for over decade now, i've a fan since i was in elementary school, and i just wanted to say thank you for everything. I've been feeling not so good recently and this kind of light hearted content means a lot right now for me. I hope you are well and continue to be, thank you again for being on this platform, I don't think i would be who i am today without your content being such a genuine positive influence on me at such a young age. take care!
Some advice that could be helpful: When giving an AI tool commands, you have to make sure to use precise language. For instance, "turn [object] into [object]" is confusing to the AI because "turn" can mean spin, rotate, provide (e.g. "turn in"), change, etc.. It's better to use a more precise word like "replace". Hope this helps!
tip: make a rough selection and include more background instead of the precise selection. This gives the ai context of the surroundings and gives it a bit of a wiggle room to generate
I can't believe Jack is using AI to write and produce his videos after his Council died in a suspicious car crash
cant believe jack was replaced by ai after overdosing. feels like hes still here with us
there was nothing suspicious about it. just a really unfortunate accident
We live in a stimulation
I thought they'd joined the WGA strike, but him killing them makes more sense.
I like the idea that the entire council was in a car together
This was way funnier then it had any right to be, PLEASE make more
yes! there need to be a volume 2 of this. I haven't laughed this much in a while
YGS
@@tadasgame I'm just 30 years old and do I really have to say it?
back in my days this would be considered shit posting, I've just watched 22 minutes of numb dumbing content.
"make me fat "haha here is a stock photo of a fat kid with weird edges that even an newb photoshoper wouldn't add "epic! XD"
I had to stop it at 22:37 this is seriously what makes you laugh?
"The grape lady" was funnier then this whole video?
I'm pretty sure that even Jack would start laughing just by hearing the sound of that ancient video. I could sample 5,10,20, 30 minutes of this video and nobody would laugh at just the audio.
"I haven't laughed this much in a while"
that's just depressing
@@svampebob007 I'm 34 years old and I'm sure I have the same exposure to the internet history as you, if not maybe a little bit more for having been around 4 years longer. Bro, what the hell are you saying? Let people enjoy what they enjoy, that's what the internet is about. Shitposting doesn't exist in objectivity. Most of GenZ content is "shitposting" to me, because I don't get it/enjoy it. Most of contemporaneous art is also shit for me, because I don't get it. But the universe doesn't revolve around me and my preferences, so I know not to judge what other people enjoy based on my own preferences. Unnecessary, man, totally unnecessary. Keep the judgement to yourself.
@@svampebob007Cry about it + no one cares
The "teeth" prompt generating toothbrushes is kinda interesting, because it reveals something interesting about he dataset the AI was trained on. Within the Data set, probably most of the pictures with people in them also had teeth in them, but it's not something you, as a person, would think to tag and point out to the AI. However, on pictures that have to do with oral hygiene or dentistry, you would feel inclined to add "Teeth" to the pictures tags. In the training process the AI then looks at those data sets and concludes that the difference between pictures with and without "teeth" is stuff like toothbrushes and wide open mouths. The data set had a human perception bias, almost like a blind spot, and as a result the AI doesn't know what "teeth" are.
That’s actually really interesting, I wonder if adding images of just open mouths tagged with teeth would fix the problem?
@@hepatitisf7495 mind you i don't train AI or anything even resembling it, but i'd imagine you'd have to go through and tag all images with teeth in them as teeth, otherwise the AI would start assuming that open mouths or other parts of the mouth are teeth. i'm guessing it needs some of that outside consistency otherwise it'll confuse itself
lol nerd
they could pull a feather out of a certain sodium websites hat and imply tags based off of more general things and maybe also add a system to remove certain implications if they arent in the image
The photoshop image diffusion model is quite terrible compared to what's out there (and I image the problem of overfitting that you mention could be solved with, in order of descending difficulty: 1. a superior image-to-text model during the training phase; 2. Adopting "Low Rank Adaptation" techniques to fine-tune the model for specific problem areas. This method would avoid the need for retraining the entire neural network from the beginning; or 3. even just by using negative prompting to further refine the output).
It is true though that what you're referring to is widely acknowledged as a big problem with the 'big data' strategy modern AI research and development employs.
However, it's been said that the future of AI may well lie wuth the 'small data' approach, i.e. quality over quantity. With this approach researchers might, for example, use small, highly curated datasets to finetune a larger model such that it's broad knowledge can be transferred to something precise and accurate, eventually creating a conceptually perfect text-to-image model in time.
Although we first need to learn how to crawl before we can walk.
Jack:"make him look angry".
Jarvis:"Hispanic it is".
is that racist? it looks racist
That has to be some flavor of racism
@@robloxperson8579 why lol
It's almost as if GIGO applies to data sets and if the data inputs that it learns with are garbage and not carefully vetted, then guess what the results are going to be??? Haha
@stephaniegalliart859 that's a harsh conclusion based on ONE case, but it could be true
although "angry hispanic" isnt really a stereotype ive ever seen
Some explanations of how Photoshop Generative Fill works that he must've missed:
1. The image completely re-generates whatever is within the selection. Imagine it as the AI "deleting" whatever is in your selection, then re-generating it on it's own. This is why when he selected himself, it completely remade him as a different person. Or why when he selected the whole screen, it generated a whole new image.
2. Correct me if I'm wrong, but I believe that the "prompt" you type before generating an image is what you want the image to be of, not a command that it follows. Instead of writing "make him smile", write "smiling mouth", because you want it to generate a smiling mouth inside the selected area.
3. It doesn't understand logic behind the words written in the prompt, it only sees the keywords themselves. This is why when you for example wrote "make my arms into spaghetti", all that it sees is the word "arms" and "spaghetti", making it generate arms holding spaghetti rather than the arms themselves being spaghetti. And when you wrote "no shirt", it just sees two separate words "no" and "shirt", and since it doesn't know what an image of "no" looks like, it just generates a shirt.
Thank you as someone who understands this stuff I was going crazy on the way he says stuff like "/add/ chicken" 😂
To be fair he's using it like you would chat gpt so it's a fair mistake.
My digital media teacher told me about this so I already know how Generative AI works.
(Well, not everyone though.)
But didn't it understand "remove can"? Maybe this AI was trained with prompts that can represent actions, aside from objects, so to say. It might not understand "convert my arms into spaghetti", but it could get simple prompts more typical of photoshopping, like "remove thing" or "make thing blue". But IDK I haven't watched the whole thing yet, just some speculation, or my thoughts on how this tool would be more useful.
@@jonpatchmodular I wish it did work like that but sadly it doesn't. Like I mentioned, the AI has no idea what is inside your selection (point 1), so if you select a can and write nothing, it's still going to remove the can because it has no idea there used to be a can there. For the same reason "Make thing blue" doesn't work, because it can only generate inside the selection, and doesn't know what used to exist in the selection. For this reason, "blue can" would be a better prompt than "make this can become blue"(although that still probably works because it has the words "blue" and "can" in it)
EDIT:
Photoshopped has a built in feature where if you write something like "no can" then it automatically makes "can" a negative prompt, meaning that it'll try and avoid it. If your selection is completely covering the can, this is still uneccessary.
I can give you an image of "no". God.
Haha, buuuurrrrrrn.
Watching Jack fucking around with stupid AI tools in real time is way funnier than it has any right to be. I am so down for more of this!
" is way funnier than it has any right to be"
Does it though?
"make me fat"
AI: crudely Photoshop fat kid in that looks nothing like him at all, just some AI fat kid
and then 22:37 (pause it right there is that funny?)
Forced comedy will never beat just comedy gold:
"dutch reporter falls off boat into water"
@@svampebob007 ruclips.net/video/So1X4BN03jk/видео.htmlsi=umB6QMVsoEKtCOPV
@@svampebob007Holy shit we get it dude your comment essay isn't going to change anyone's mind on this being funny it just makes you look sad.
@@svampebob007if you don’t like it.. then don’t watch it? Who pissed in your cornflakes man.
@@svampebob007 okay, we get it, grandpa. youre offended. comedy was better "back in your day" or whatever. is that what you wanted to hear? no one forced you to watch this. if you dont like it, just leave a dislike or something and leave. no need to leave multiple comments berating people for their sense of humor and acting like some kind of comedy elitist freak lmao 💀
Jack has taken making others make his content for him to whole new dimension. Now his audience provides prompts for the ai to generate it.
Jack has always been an AI this whole time we are its willing participants
Yeah, makes you wonder why people even watch him. He barely does anything
"This is the future that liberals want"
- Hen Shapiro, 2023.
Take my like and f*** off
Breaking Bad photoshop
Robot Named Finger
jesse we need to cook
Waltuh, stop removing me from the photo with ai waltuh
@@simpIeidiotWe were destined to be together and you know it waltuh
Jesse, stop turning me into a woke chicken with ai jesse
I love the whacky characters that always pop up in the Jacksfilms submissions. Diarrhoea Megan, Shower Shitter, Wario’s Nipples, classic characters
truly the recurring characters of all time
did diarrhea megan show up? Didn't catch her return.
@@lisyaorancrazed no I’m just listing some of my favs from through the years
i love you too (i am ashowershitter)
@@ashowershitter omg THE ashowershitter! I’m your biggest fan
11:30 literally the best machine learning generated hands ive ever seen
I was looking for this comment!
Frrrr
22:08
The new self OC looks great, Jack. Can’t wait to see him in artfight.
I absolutely love the whole AI aesthetic, how it doesn't really understand implied lines, legs, fingers, etc. We see a lot of really creepy undertones in the image where you had the AI fill in the rest of the room. It's like a room, but everything is subtly wrong. There are what appears to be door handles on solid walls, or "box shaped" items that the AI knows we all have, but is severely confused by.
Looking at AI images is so reminiscent of ketamine or other disassociatives in a way. Things are a bit alien.
Deepdream turned some of my pictures into something that seems like it can simulate psychedelics.
The morphing... the organic look that edges can get. The swirls... the spirals. If you aren't familiar with DD and you love this type of trippy stuff you should definitely check it out.
yeah man earlier AI stuff had some serious kitty flip vibes
"that chicken absolutely has pronouns in their bio" 💀💀💀💀💀💀
The ai turning Jack into a Black Woman is my favorite running gag
20:43 Jack, you really need to warn before jump scares like that
Love to see that jackisanerd 2.0 is already off to a roaring fast start.
Randomized pizza tower part 3 wheb?
@@OkthenBi Never.
@@KUR0.KALAM1TYrandomized pizza tower part 3
With the "HAPPY" prompt I think the AI thought that since in many photos of happy people there is confetti and other stuff in the air (covering the face) that it should add that to your face to make it happy
Jacks extreme joy doing this is so wholesome and infectious. King of YT
This video had such great energy, pure positivity and jokes but very relaxed. I laughed out loud multiple times, thank you mr. film!
2:54 why is lisa moaning jack?
For future, I think it’s better if you understand what the AI is actually doing with your prompt. It is a replacement tool, not an edit tool, it does not get given anything inside your selection and the text promot is a description not an instruction, so if what’s outside your selection does not include necessary context (like skin tone, shirt color, beard) it will choose for itself. This also explains why stuff like “remove X” confuses it, because if you select X then it doesn’t even realise X ever even existed so you have to decide if X would be expected given the context; if it would be, you tell it to not include that (such as “no mouth”, since you’d expect a mouth to be there), otherwise just don’t give a text prompt then it will remove anything “unexpected”.
But... the very first one he did was "remove can." And it did. It removed the can and replaced it with what the shelf (or whatever was on the shelf) might have looked like underneath.
@@KelpTheGreat It would have also done that if he typed nothing. Can’t remember where, but there was definitely other times in the video where that *didn’t* work and he got frustrated, so you’re just cherry-picking
Tears are literally streaming down my face and my sides are in pain from laughter. I downloaded the beta and am going to do this with my friends tonight.
my husband had to take his headphones off from the other room and ask if i was ok because i was laughing so hard for the entire video lmfao
Excuse me, but may I know which version of Photoshop has this AI tool?
@@Pico_CaptainIt should be the latest one, I think. If not, you might have to install some experimental branch or something like that.
16:25 Pov: Jack made direct eye contact with you (he has the ability to turn anyone he looks at into cheeto dust)
It’s Medusa, but instead of snakes it’s just really long cheetos
8:12 well… it’s a writer’s strike, you really think they’re about to fall for writing on their signs?
I was absolutely dying at the cheeto fingers, as well as pretty much the entire video. PLEASE we need a part 2.
24:56 to be fair, I can only imagine that's exactly what an average Italian plumber looks like.
“Reality is often disappointing”
-Purple chin guy from Marble
18:10 this dog is fucked up bruh
quit i’m gonna pee in my bed
14:35
"i love this band, though. i've been a fan of liuexxxu. they're a swedish death metal band."
this was so funny to me for some reason 😭
Somebody needs to make a compilation of Jack's laughs. They come in a rainbow of variety from sensible chuckle sll the way into creepy clown territory
0:51 didnt know it was possible for a human being to produce such an indescribable laugh
god damn jumped scared me
that's what woke chicken does to ya
i'm pretty sure it's pitch shifted but it's still really funny
Geese need HUGS
5:20 new dream design??
I knew someone would say it.
I wear mask with smile
The fusion of Dream and Kwite
Kwite
no, thats MF DOOM
18:50 Hehe hampter
I would love to see you revisit older yiay topics but with these new ai tools, like the fix your flags of fix your twitter bios
22:36 the clown on the right looks like the polish flag 💀💀
Nothing has ever made me feel more terminally online then when Jack said "What the Ratboy Genius"
Idk why, the construction worker bit was funnier than it should be
the orange cheetos dust one was my favorite because jack looks like a mildly disgruntled office worker having cheetos dust flung at him by someone off screen
22:18 Jack's official OC lmfao
4:46 Giant rodent timestamp
i didnt see the dog walk in and only saw chip laying down and i fucking thought it was a PIG..... 😂
2:06 why did the AI turn him into a Garten Of Banban character
Channel of jackjack
I THOUGHT THAT TOO 💀💀
ZUMBO ZAUCE
CREEPY GREEN???
He looks like a chinese bootleg figure toy of cweepie gween
19:22 tomska is that you
I literally said that before checking the comment section
28:13 Jack sees "Great personality", I just see "Serial killer"
0:05 right off the bat i can already give a tip. photoshop's new ai doesnt accurately produce images from instruction such as "remove this object" its much better to describe what is in the background so the ai knows what you want it to replace the pixels with. "laptop with stickers" or something like that is probably ideal for that scenario
Yeah he used the ai completely wrong
@@incriptionI don't think the point of this exercise was to use the software correctly lol
But it worked tho? It removed the can?
@@Zeffer32 well yes but im just pointing out it’s not the intended way of using the photoshop ai feature.
Also there’s lots of points in the video where the ai makes something entirely different to what he expects because he used descriptive language in the prompt instead of telling the ai exactly what he wants to be generated
Adobe is evil and generative ai is theft
You better have pirated PS.
Why?
@@joaquimrietveld8297 Because fuck adobe
@@joaquimrietveld8297 because they make it cost like $30 a month
25:15, thank you for the feature council, and jack for the wonderful images. i am ashowershitter. thank you.
World Famous ashowershitter! Congrats on the feature!
@@bbbeththedorkyass i am world famous fr, i shit in everyone's showers
ladies and gentlemen, this is what techbros are trusting the future with
7:42 looks like the facial reconstruction of a unknown crime victim oh god
25:33 What's number three, John? WHAT'S NUMBER THREE?
Oil, duh
@@winzigerflashendeckel6894combination of 1 and 2 ig
25:44 You don't know what a number 3 is? Your clearly not from earth.
Puking
13:56 this looks like a Muk bang/apology video.
sband dhbhdydhhhhhahahhahahahahaahAHAHAHAHA
8:52 jacksfilms making a ratboygenius reference is a very welcomed surprise! :3 glad the series isnt forgotten despite the lack of uploads.
19:15 looks like a mix between Johnny RaZeR and TomSka lmao
14:45 jack finally acomplished his dream of working of lighting fast VCR repair with the redlettermedia guys
15:28 looks like jack is just trying to take a photo and some random kid ruined it with their cheeto fingers
15:00 just a very silly happy boy 🤩
21:10 Am i the only one who sees a bunny with wings😮
My lungs hurt from laughing, thank you Jack
Thank you Lisa
5:03 I love how genuinely hurt and mad he sounds
19:16 “make him funny”
Well, AI got it right, it gave him TomSka
the reveal at 3:30 caused me to do an actual spit take
6:40 THALASIN EXTENDED EMOTIONS!
NARGE
16:35 POV jack film watches you as you disintegrate into cheeto dust
I wasn't expecting the Onision jumpscares
hi jack, i've been watching your content for over decade now, i've a fan since i was in elementary school, and i just wanted to say thank you for everything. I've been feeling not so good recently and this kind of light hearted content means a lot right now for me. I hope you are well and continue to be, thank you again for being on this platform, I don't think i would be who i am today without your content being such a genuine positive influence on me at such a young age. take care!
2:17 “hEY….. that’s pretty good” rent free in my head now
(24:50) To be fair, the AI is clever. She could be Italian and working as a plumber. She's just on her day off.
7:57 is just Elon Musk before his hair transplant
Jack when Sssniperwolf is outside his house: 6:41
Jack when Sssniperwolf “apologizes”: 7:01
Jack when SSSniperwolf finally stops stealing content: 9:49
Amazing how the AI removed Erin when you requested it to "remove beard"
horrendously underrated
26:21 Best joke.
Literally dying the whole time!! I would legit take this as a regular series!!!
Photoshop videos always make me giggle so much , more please john, I love it
18:03 I know tobuscus aint all that great these days but its nice to see Gryphon again
4:17 the ai put that painting there because of the green on your hair from the greenscreen. Thats crazy.
11:20
Level 1 trainer pads
vs
Level 99 hands of the beholder
7:25 BanBan VIII looks neato
Jack you HAVE to make more this was TOO GOOD
25:06 I was just thinking that looks like Steven Ogg!!
I absolutely love these photoshop AI videos. They're always so entertaining
Some advice that could be helpful:
When giving an AI tool commands, you have to make sure to use precise language. For instance, "turn [object] into [object]" is confusing to the AI because "turn" can mean spin, rotate, provide (e.g. "turn in"), change, etc.. It's better to use a more precise word like "replace".
Hope this helps!
I know it wasn't long ago but I miss when this kind of stuff was all that AI art was. It's fun seeing what an algorithm thinks a furry or a minion is.
Please make another one of these. I laughed more than I thought I would.
2:53 the way he just carries on like he didn't say what he said makes it so much funnier
25:55 shout out to this geniua
tip: make a rough selection and include more background instead of the precise selection. This gives the ai context of the surroundings and gives it a bit of a wiggle room to generate
29:53 Who’s getting the best head Ytber addition!!
1.) Ai LazarBeam
2.) Ai Markiplier
Or
3.) Ai JShlatt
I loved this Jack, please make more!
1:38 weird how they have them but still say they're against giving them to you. They just like to keep them :3
12:08 this reminds me of that image of a hand holding a gun pointed towards the viewer.
You have to admit, typing "funny" resulted in one of Jack's biggest laughs. Well played AI.
12:46 considering the size and shape of the spaghetti, I’m a little afraid to guess what that one pulled from
The big boy at 10:58 is doing the gordon ramsey pose
Dude jack you’ll always be funny. “itS gIVinG” 😂😂
i love your laugh, particularly the way it blows out my speakers
Honestly one of my favorite videos I’ve seen all week
7:01 looks like if jacksfilms and pewdiepie had a baby
i died choking laughing on this video. im glad jack is returning to actual funny content
6:11 AI is racist, all angry men have a darker skin tone for some reason.
Bc when you yell you go red
Touch grass
@@404TVfr get a job.
@@404TVfrhi guy I agree
5:14 the ai really turned bro into club penguin 💀💀💀💀💀💀💀💀
Jack just straight up created Tomska with that funny prompt