#general
1 messages Ā· Page 156 of 1
it feels like I'm going through withdrawals
but its novembor? š² š² š² š² š² š² š² š²
I think they'll probably release it in december as they always had with other models
i know
but I really hope the preview at least comes out early
nah.. it's novemeber. December will be stable version on APIs for enterprise use
and it's as good as lithium or orion was
lithium and orion are both gemini 3,
still preview
technically
WIP
How do I access the new code arena
it's not great.. sometimes works great but most of the time meh
Very nice, itās wrapped but thatās good enough almost perfect
thnx, but wrapped?
jesus christ why are there so many nondirect chat ais
i got 3 non direct ai in a row bro
Yeahs itās a technique they used to prevent copyright
yeah lol
is ring 1t underrated
Didn't you see my previous message
.
wrappingā is sometimes used to describe the process of modifying, transforming, or re-encoding video content so it is less directly recognizable and therefore less likely to trigger copyright detection.
Itās not about magically avoiding copyright, but about making videos appear ādifferent enoughā to automated systems (like YouTubeās Content ID) that they donāt match the original copyrighted source directly.
Anyone?
I'm just saying that's what I was told by others
It's from baidu
baldi?
oh baidu
qwen3-max-2025-10-20 exists what
what
wait
wdym
"im claude from anthropic" no youre not??
Is there any old conversations
yeah
That's why
Man it really sucks huh?š¤
What it
How broken and utterly unreliable hallucinating ai products really are.
You said or typed "unreliable"
I donāt understand what you mean my apologies oops I think I messaged the wrong person I was talking to Janix
It's meaningless since it's just a way ai to escape their current generation process
they can't find any other reliable data in their training set
Itās nuts and itās crazy that they charge some of these places and companies charge a lot of money
@ them
But yeah, well you said it was 100% correct
Even the generations suck it still consumes computation powers
Itās wild., on top of that the extreme censorship and model personality building and the way theyāre just designed is just frustrating to be honest and thatās the best word I could put
I mean, it just shows the world we live in. People are so sick of Google and Google search that theyād rather trust and use unreliable, hallucinating, factually, incorrect sometimes and all the other issues that come with AI rather than. And I canāt blame them.
How Corporate Greed Is Killing The Tech Industry
A 2025 study revealed extreme CEO-worker pay disparities in tech, with CEOs earning in a year what employees would make in centuries. Since 2019, over 800,000 tech jobs were cut globally, while profits grew 40%, often justified by AI or operational efficiency.
Companies like BlackBerry, Google,...
You know thatās why if it is a bubble and white a lot of these companies arenāt able to make money
Itās paradoxical
Are you good at prompt engineering?
"Yes we need a lot of companies competing with each other to reach agi"
true
"hi im chatgpt" and its claude
bruh, clear you chat history if that's not the thing you wanted
Shame on these companies and people who know better and still try to profit off the ignorance and mis informed public who are now reliant on these tools
I was reading a post yesterday on Reddit about some guy who ran up like 13 K api bill with Gemini
He says, and I quote āoutstanding bill for $13,371.
This amount is not just something I can't pay; it's a sum that completely ruins my life. To put this in perspective, I am a student in Bolivia, where the average daily income is around $4 USD. This amount is more than my entire family could earn in a decade.
ā
tha can just.. happen???
yeah no screw this how can i cancel my 300 dollar credit
What, that's google cloud n not google gemini related
oh wait
man i was trying to get a paid gemini key not witness some guy get charged 13K dollars
I heard even more stories on Reddit people getting charged thousands of dollars dude over 20k
by paid i mean
gemini 2.5 pro.. smth idk bro
im not a pro with these api keys
these are useful
Itās not a one time thing
Itās crazy dude
Not for me
Ty
Yes but I haven't gotten it even when I tried
bruh that takes like 20-30mins to get one output from Gemini 3.0
and u waste alot of credits using 2.5 pro
Google just needs to release the damn thing, people are getting crazy
instead of wasting alot of power running 2.5 pro to get one 3.0 pro output, they should've just release 3.0 pro preview
i got one in the second attempt tho and another one in another account in like 3 mins
but yeah i agree
it sucks having to do it
not worth it for sure
I canāt even share the prompt lmao
Fr?
See this is a different kinda of bias, a completely positive message no harm uplifting and blocked.
Hi Rich here I am a dinosaur hat created award winning video's and taught social media in the early 2000's
I think 100%. Too good results for 2.5
this might be the newest fad like how we have sora
this is like realtime video editing
very surreal
@echo aurora sorry for the ping,is there a way to dowload my chats?or request my chats data?
you can try it at odyssey.ml but it's not great tbh
the quality is as good as the first will smith eating spaghetti
so it'll catch up to where sora 2 is now in 3 years maybe š

You do know they all have the same training data for the most part
are you talking to me
The only main difference isnāt proprietary data and how they train
Thanks, I've removed. Don't hesitate to ping our @ Moderators next time though as they can help as well.
But the data sets for the most part are pretty much all the same
oh thats a pingable role?
nice
ty
Guys is perplexcity comets agent thing working? Im asking it if it can take over my screen or even open a simple tab and its saying no you'll have to do it manually or something
There isn't' a feature in app that lets you download all chats; however, you can reach out to privacy @ lmarena.ai for data requests.
It is 
Yeah it's great
Guys, Iām really conflicted about artificial intelligence to be honest. Like the lazy in me loves it, the care free version of me but on the other hand, it comes with the cost of some real ethical dilemma is here lol
Why does it have to be like this?
Well you got spoken to like a bot because u messaged like a bot
U know tryna break it down to ur level
š¤£
Spoken to like a bot? What is that like a secondhand citizen? I never heard that thatās funny as hell
hey how often do rankings get updated?
@drifting crow
king von
I got a question
what
lmarena.ai or google ai studio
hello
grok4 sucks
we need grok5
(and gemini3)
yeah it sucks
but i like the UI
grok sucks lol
cant listen to audio
unlike gemini which was released way before
You can do so many ā¦interesting things with Gemini
guys why isnt comet working?
Why the models in lmarena take long time to provide answer??
wdym, grok's website has been around for a while
depends on the model, some of them have long reasoning times
Can you elaborate a bit more? How long have you been waiting? Is it all models?
Do yall have unblocked ai to use in school?
The scene quality is impressive but if only the model was smart enough to compose the scene properly.
Iam enjoying this site very much, at the same time worrying will this service available always? For free or any payment plan in future?
hello
Yup
7 MIND-BLOWING Things About Japan That Will Make You Want to Book a Flight TODAY! šāØ
Ever wondered what makes Japan so incredibly unique? From bullet trains that are NEVER late to vending machines that sell literally EVERYTHING, Japan is a country where ancient tradition meets mind-bending innovation!
In this episode of Wonder Planet, we're...
/sora 2
no
it's completely free for now to everyone (we're deciding to make it a paid service tho in a few months)
lol
fr
Huliou 2.3 just dropped
Go check it out 4 free tries (:
no sign of it on lmarena yet?
Any vison model basically
on LmArena?
i couldn't find any lol
battle works?
Why wouldnāt it?
how about for batch processing
i js wanted to chat with the model directly
the vision models refused the image I gave it
Not so simple
ik
Limted
Theyāre all pretty much the same dude it doesnāt matter. Thereās really hardly any difference from one to the other.
really is a shame
yeah, fair enough
Guys why is my LMArena is generating infinitely
are you using the coding feature
What are you trying to analyze?
No
i'm trying to run 21 pages to create a study guide
wanted to see if it can handle it
I think what she means is her little thing is spinning, but nothing is generating
Use lm notebook
Or use m2 agent itās free currently
Pretty decent I like it
What model?
Gemini is a strange model in my opinion for many reasons
it works well
i ground it with google search
Yeah people say that, and it does have those perks you are right
3.0 will drop soon
I just feel like itās not as intuitive as ChatGPT though
Yeah, thatās nothing to be proud of because ChatGPT is a pain in the ass to use sometimes
But the numbers donāt lie
i get that, but it works just as well
sometimes better
No, I hear ya and teach their own
iām waiting for 3.0 to drop
Iām just saying I just find it kind of a strange model because itās in second place, but it doesnāt have the numbers that ChatGPT does
A lot of people are. Iām a bit skeptical
Ik itās quite weird, but itās what the people like
even if the usage is significantly different
Ik, itās tragic. However chatgpt does have some polished features, like analysis and the ability to create PDFs and graphs
Google has only 12% of users that ChatGPT has and what makes it even more strange is Google is offering so many different AI products itās crazy.
OpenAI is greedy
what can I say
Google no bettterš¤£
And google shouldāve been in first place
They had a Headstart and AI development lol
They had the money to talent, and I donāt know what the hell happened
I mean technically you can use there premium models for free
without LM arena
It sucks what you have to do and what you have to give away just to stay in the race
I mean yeah, whatād you expect. Deepseek bursted the ai bubble
and how we people saw things
Fr open ai greedy asf
mhm
Actually, I take that back. They canāt be greedy.
Theyāre not generating any real money
š¤£
They made like 22billion and had like 9 billion in expenses
They are expected to loose 39 billion by 2029
Worth more then space x wtf š
Bigger then TikTok lol
The leaderboard would be a good place to check this
right, but I can't even use the vision models directly
is that on purpose
Sound is not generating in my bot
There are times where some models only sampled in Battled in only, but overall should be able to in text modality
Not all generations are going to have sound. Itās random if the models youāre sampled have sound or not.
it didn't let though
Today's 1st video to last nothing any sound of 10 videos
U can increase ur chances with proper promoting
I was tested some types of prompt but it's not working how can I understand something
What are you trying to Genrate.?
Iāll show u
You got 50% odds of hitting it and got 25% of hitting 2 with none
@keen beacon iam generating image to video Pixar style 3d cartoon yesterday it gives sound as I wish but today nothing
Ok show me what image Iāll prompt it
Can u make sure it
What kinda of sound?
Music and some dilouges
Ok
What is problem did u know
I donāt believe thatās true
And you explain a bit more? Are you running into an error
yo
so basically I go into the direct chat mode, use a vision model. Attempt to upload the image and it says that some models don't support this type of file upload
it was a png
Iām having my buddy tester right now by generating
We will see if itās two for two
What is this for
It's good can u know what is problem to generate sound
How many languages do u know what is ur name
Country iam from Indian
Thatās awesome man. Thereās a lot of cool Indian people Iāve met. Itās a pleasure to meet you.
Which model is this?
2.5 pro
Agent lol what is it and how @keen beacon
You can try it free with mini max
Build with next-gen AGI: MiniMax offers Foundation AI models (text/audio/image/video). Power products like Talkie, Hailuo AI & enterprise APIs. Scalable, secure AI solutions for 50K+ global businesses.
It can generate video audio
It can build stuff
You donāt have to install or download anything
You could do batch download batch prompting pretty much do anything with this damn thing lol
I know I can't try for free with minimax
It can make more special of lm arena or best i don't
Yes u can
Itās free for everybody right now for a week
Yall are missing out lol š
the api is free too if you are interested in that (for m2)
Itās really fun and itās really good actually
Oh get me a link
I never vibed code in my life. First time I ever did it today lol
It can build backend front end
Iām really enjoying every half
Can have u tested something special show me
Bro just try it lol
How
I love hailuo
Build with next-gen AGI: MiniMax offers Foundation AI models (text/audio/image/video). Power products like Talkie, Hailuo AI & enterprise APIs. Scalable, secure AI solutions for 50K+ global businesses.
Same
There awesome
I would love to do the content partnership program. I donāt wanna be compromised. lol
I donāt wanna shill š
For test do u know which ai is do this think
Let me think think first a second
That looks like google image size
But the tears and audio off imo idk š¤·āāļø
S That is txt to video veo3 google try veo 3
Oh
I suspect thatās what it was, but the only reason I thought that was because of the video resolution aspect ratio
Google the only one that does true 16:9
No edited aspect ratio google gives water mark our of video so i cropped it
Yes it is free with water mark try veo3 it gives 20 credits per google account u can use so many accounts and try 10 credit 1vid
And having another thing of google it is also free
image to video
U can see the water mark is out of video u can cut it
Nice but why this
And i send u a 2 video of wan and veo u tell me which is best
Ok sure
U can remove water mark of sora
Which is best
Are u a creator or youtuber
Some thing what u do
Veo is smooth but the sound is i can't say but both is in same use
No I donāt do social media stuff
Maybe discord
I got nothing worth while sharing or any good ideas lol
Some thing special video which u created best ever and u love
My sister is a blogger
Oh great
I guess
How much time means from eyes 5yrs blogger
Itās all ad money
I donāt like ads I donāt like seo
I donāt believe people should be treated that way
People want information not be bombard with some crap
I can't understand what money and adds what do people
Itās not a big deal on its own
But me personally i never buy anything if itās advertised lol
Idc what it is
I used to do door to door sales lol
hello
Sorry for that
Nothing to be sorry for, I live in the biggest shopping mall in the world
They even got me in the car on the radio first thing in the morning driving to work some ad Iām not even awake yet and somebodyās already trying to sell me something š
Last time when u came india
Can u talk in another language
My mindset is that if someone has to sell me something, I automatically think, āWhy are you trying so hard to sell it?ā If it were really that good, it would sell itself.
I speak the international language of being human
They understand that worldwide even the deaf and blind I understand the language I speak
Madschode
š
Kya hi ye
Believe it or not I grew up watching a lot of Indian movies when I was younger
The dancing singing action
Ri tum ise kuch to samajh te hoge
I don't know who is in video and iam so young
I think the man is Mithun chakravarthy
How much is ChatGPT plus in India
Thatās probably from the 70s or something. I donāt know that was way before my time.
400rs per month it means a 7 to 9 usd
Bucks means
And i never pay to any ai tool
Yeah you and 98% of ai users lol
U know how rich open ai would be if everyone paid?
Only 3% of users pay for ai
What does it do?
Google flow ai free it gives 100credits per month 20 credits per video ,txt to video image to video frames also u can use so many g mails
It helps to define Claude according to your liking. As the normal claude modififes code in ways that you dont want.
Give it a try
Oh cool. Is it like a recent development? Thatās why itās taken by storm or?
I never used Google veo3 or flow
Iām not a coder š more gen ai stuff media images lol
Why it,s cool and magic
Oh cool man
It is solving key issues no one has
Is that you who got that award?
No, not me but my buddy
Oh well, I just wanted to say congratulations buddy. Thatās awesome.
Hi, just to say hello, just learning all I can about AI.
Hi
How can we get it featured in here?
Yep, how can i get it to be featured is what i am trying to do. It solves key issues
Just look at all the components:
I did. I am just wondering if there is a way to get it to the right eyes to make it happen
Wow
hello
<@&1349916362595635286>, also in #ai-news , #codename-discussion , and #share-prompts
why does it feel like 99% of the video generations arent being voted on
just people joining on alts to create slop wasting electricity
There are many models that have been added weeks or even months ago but never got listed to the leaderboard, e.g. Qwen3 Omni 30B, is there a bug?
Pls new system prompt š
#video-arena-2 message <@&1349916362595635286>
Someone pls test grok heavy
because they dont care about vote they want free videos
exactly
they need to fix people just creating slop because i feel like this might ruin it for everyone if ai companies stop providing free access if they arent getting the data they want out of this
how is the limit of the chat is decided before "something wen wrong" happens
How to fix forever generating Claude 4 5 sonnet thinking
Hello
Hello guys can i ask something please about prompt?
Apparently, we need to wait until december, for a decent coding AI..
U canāt itās a cap it goes away like every 24 hours? (Maybe)
What i so to come back sound music etc in my videos
What do u mean?
Means having problem to generate sound in my videos isaid
did they remove lithiumflow
What is it
rumors say its gemini 3
Itās all random my guy
Use grok
Itās free comes with sound
That feeling when nano banana craps out bad gens, and you gotta wait an hour because they made their limits more strict
No
I will try
Nano is so annoying
i create the vedio whre i csna find
I've been trying to use it to create discord stickers of characters
So far very succesful, but it takes so long because of limits
Nano is lazy and stubborn
@keen beacon
@cold oracle
How to generate multiple videos it is allowing only 5 per day if we delete genration will I be able to generate again pls help...
this thing also works on claude
nope
even if you delete you still cant generate a video
Oh thank you
Could you recommend any other free ai tools through which we can generate by prompt
sora ai is the most popular but paid
veo 3.1 is also good but paid also
that one grok tool idk what its called
Oh so no free resources I guess
this is 4.5?
yes
"they" could mean anything, right? so yes or no could be both valid
If you follow the fuzz online, Sora is most talked about.
If we go with the number of paying users SeaArt is the single largest without any close competitor.
But below we find Hailou. WAN and indeed Veo.
Grok is free
It just isnt as good,.but you can chexk it out
Thank you for the reports. This has been actioned.
That's creative!
Indeed.
I have an idea - look here.
For my story I made a number of clips with an open book for each episode.
And I have a hunch that person have used a similar technique.
Nice! Itās actually a template you can download but how does he get the stylization in color pencils ?
Some of them came out silly weird - this is a fail, but a funny one.
You can download the template and upload ur own video.
Sorry, somebody asked me to help them earlier so thatās all the information I have. He just wanted to know how you get that stylization that colored pencil look?
I can only say how I would use that on SeaArt, and that is to either choose a model for it, or use Control net & line art.
What model can do this ?
Maze did have a line art setting also making like art out of photos or images.
But Maze is pay only now since last week.
It needs to be video
You could literally upload any video to the template and Iāll have that flipping effect. But how do you turn a whole video into a color pencil look? lol
Lets see, Alpha lineart 2D might work.
That's why I think it's done on local, and he have run a script with the checkpoint loaded.
Yeah could be
Lol! Mystery solved. There it is, a website specialising in that very type of videos.
Then break it apart with videoproc and batch process all frames, and then assemble them again - could do that in 15 mins.
Ya wow
how many tokens do you use per day?
2
8
Batch processing frames? Umm yus, think I done that one. <- lies Aibo do that regularly!
Reason for that is to get colours right, as they sometimes shift in one generation to the next.
SeaArt AI is imfamous for that, even when one do use the same model.
So to add one 10s clip to another, I batch process the frames instead of going deep end trying to adjust in HSL or CMYK in the video editor.
<- Lives in the shallow end of the digital gene pool. š¹
But: Macro programming I know by heart.
Has anyone encountered Neptune V6 in battle mode yet?
(it should be good, because it would be new Claude Opus)
@sullen quest thanks mate I will give it a try
No
Hailou have always been good at faces also at smooth camera movements - their prompt interpreter is quite good also.
Though one frequent user claimed he got better results when using machine translation to chinese - which cannot have been perfect.
But yes, even they tried to get me hooked with a long free period I never subscribed, DDG provide HD size for base price and also got smooth movements and good prompt understanding.
Nice, I delete clips also after showing here.
That wasnāt too bad was it I mean the quality
I get this UI accidentally. How to get it in a new chat?
Indeed, Hailou used plenty of BBC footage for their model, so I'm not surprised to see that clip.
Well, one thing itās kind of lacking it kind of abstract
Itās very rigid in a good way in a limiting way at the same time
But but, I'm mad.
If an AI do not provide 1920 * 1080 at base setting and base coins / enery / points cost I will not part from any moneis.
I guess the word Iām looking for it seems more grounded
Might as well be asking for 4K at this point š
5$ a sec
Remember I said 'base' DDG got 2 higher setting than that.
The highest one is prolly 4k - actually never tested it.
I wants 16k hdr
You're nuttier than me! šø
With surround sound that you could hear it needle drop
Crazy
I mean if youāve showed me this five years ago and told me that this was fake I wouldāve called you a liar lol
hello
So I was actually mistaken itās $.10 per second for open AI api. And 30 seconds per second with Sora pro.
i am happy to be here, looking forward to generating vieos
Welcome
This might be the only AI discord in the whole AI community that I have heard people say the same thing over and over other happy to be here. It must be a great place. lol
Here's one frame from DDG - compare with any AI you want, no mudfaces, no fuzz.
Looks rendered
@void yarrow Welcome! You can head to #1397655624103493813 for a detailed guide on how to use the bot
What was that the aesthetic you were going for?
Yep it is, and indeed it is.
My original idea was to go all out anime style, but I decided to go with a 3D rendered aestetic.
You know Iāve been thinking about that actually about the concern about having two realistic videos out there that arenāt real without some kind of some kind of label, letting people know. But by alternate and kind of giving it weird defects or non-realistic style may actually be the more responsible thing to do.
But at the same time, watermarks is not a solution
This with the exception of the protagonist AI - who is anime style. Also ads seen on buildings. This a compromise from me debating the style.
I mean, what responsibilities do creators have if the people that may be viewing arenāt as well rehearsed in AI and may believe something to be true or real when itās not without having any idea. But I guess we encountered a lot of that every day as is.
That was not a main consern of mine, but yes, it was part of my thinking. I wanted things to be one step away from reality - even though AI were not able to go fully realistic anyway when I got started.
Well, what goes around comes back
If you start putting a bunch of misleading or purposely trying to trick people with your videos
Eventually, one day youāre gonna be tricked by somebody elseās videos or even worse your own you start believing the delusion that youāve been selling to the masses.
Indeed, luckily I'm not a decision maker on such matters.
Thatās why I say heavy is the head that wears the crown
Ooops, anime figure at left holds her eyes in fear - me and anime LOL.
The anime on the viewscreen on the far left of the clip.
I canāt see I canāt open it
Too bad. Discord do not allow us to upload large videos, so I compressed it, you're lacking the codex for that file type.
That explains it, will not work on Apple.
Only on PC and android.
Sorry about that.
Oh no worries
I used to be pc guy and android guy
But hey, check this out. This is kind of odd.
Interesting behavior some these models have
H
Hi buddy
Hello bro what's app
Does the arena have a usage limit? It suddenly stops working and gives an error.
Jakub and I are going to do a livestream and answer questions today at 10:30 am pacific.
We have a lot of things to talk about--of course we will cover our new corporate structure, but we will also discuss our new goals for research, the evolution of our product offerings, an
Bye
Yes
When can I use it again?
I'm using nano banana and out of nowhere it stops working and gives an error (image) I wanted to find something unlimited
retry
Nani shouldnāt unless they limited it recently.
i think there should be no limit
Right now if you want to generate more videos for free you can try grok and mini max https://agent.minimax.io
Is there any way to use banana without limitation?
just retry
You can generate videos for free there currently in agent
maybe it will work
Ofc
But not free lol
Yes
Itās really frustrating and annoying model and itās very stubborn
Is there something free like unlimited nano, I'm a beginner and I really need it
use a diff model i guess?
Some surface does offer as low as or $10
banana?
Speaking of which that actually would be a great idea to have some kind of budget AI say under five dollars
But then again, if youāre a state-of-the-art model, youāre not gonna do that for five bucks lol
Some places offer bundles and include it for free
10 dollars is 60 reais, I hate Lula
I hear you loud and clear. I think AI is ridiculously expensive honestly and I think people, especially in developing countries or wherever even here in the states that may not have the means to a four day are getting left behind. AI was supposed to be the great equalizer
Instead, it turned into this beast that were forced to live with, itās only gonna make the gap that much more wider between the halves, and the have Nots. I consider AI luxury.
But on the other hand, they donāt give a lot of stuff away for free, but if Iām being honest, I donāt think theyāre doing it from the bottom of their heart lol
There's several close to that, but not with free generation.
Base level on SeaArt is 2,7ā¬. WAN is 6ā¬, DDG is 7,7ā¬
Oh there are but not the official channels lol
Er? What's an offical channel in your view? I mean that's the price they state on their webpages!
Where there is demand thereās always gonna be supply
Leg just say if I was a consumer, I would verify on the secondhand market to make sure that thatās exactly what itās worth to doublecheck to make sure that Iām getting exactly what Iām paying for and if the values are there
For example, if I go to T-Mobile today and I go buy myself a brand new iPhone
Brand new in the box and then I go try to resell it. I wonāt be able to sell it for what I paid for it.
Even though itās brand new, Iām still in the box lol
Oh yes, no quality control on AI generations - if you get fails on 10 generations you will not get anything back an most of them.
Surprise: On DDG you do! And that's why it's my main generator - beside the HQ size and quality.
Somebody has credit somebody needs a phone somebody has cash. Somebody has the credit but doesnāt have the cash and needs and can get the phone. Somebody wants the phone but doesnāt have the credit but has the cash.
If you get my drift
But I actually did get 2000 points back on SeaArt, but it took quite some arguing , showing generation id's and images and clips with an official before they yielded. šø
The guy whoās got the cash isnāt gonna pay full price for it. The guy thatās getting on credit needs the cash so heās gonna be making the monthly payments but he gets upfront cash but minus the phone.
So even though the iPhone could be posted for $2000 on their website and even their stores
Somebody out thereās gonna buy one for 800 bucks brand new
That is to say are you really getting $2000 worth of value from a new phone probably not because on the secondhand market itās gonna go for less
Same with all the services for ai
Always like that, Paypal have plugged a nifty trick I have used for years.
We need to ident ourselves now. Drat! Cannot claim to be from a country of my choice.
People just donāt think like that, like has anybody out here ever stopped and thought to themselves how much is their ChatGPT account worth?
Or any of their subscriptions like whatās the actual value like if you were to resell it?
How much would somebody be willing to pay for it?
And people would be shocked to find out exactly people are willing to pay and what some of these services are worth lol
Horribly overpriced example you got anyway, my new phone was 270ā¬. And it's Google free!
[Nothing phone Mk2]
Iām just saying the thing with technology and this has always been technological thing. Technology always gets cheap cheaper.
And is not valued as much as it used to be not everything of course thereās collectibles and you know things of this nature
But for the most part, think about it CD players tape players VHSās DVD players
TVs for that matter old computer, computers, old phones, old cars
It goes obsolete really quick.
Yep you said it as I were typing. All of that on that to the scrap heap already.
Well, you donāt have to look look at this place. People are having anxiety attacks. Canāt wait for Gemini three to come out.
When it wasnāt too long ago, that Gemini 2.5 came out
<- Have not turned on the TV for a year now.
And look how people are talking about it now like itās trash old š
One day, people are gonna be talking about veo3 there will be a time in our lives that people are gonna be saying veo3 is trash
Nobody will use it even if they give it away for free
Indeed, I stated which AI is the largest right now - will they be that major a year from now? Prolly not.
Will they even exist 5 year from now, perhaps not.
Some of them might be in maintainance mode only - for die hard enthusiasts.
See the biggest question people ought to ask themselves is how do you sell somebody something thatās worth pennies maybe dollars at best for 200 bucks a month and to who do you sell it to?
I mean, besides the stuff you generate the stuff, I generate the art, the images, the videos, all of it all worthless in the end if weāre being honest, itās only valuable to us but not in the monetary sense
Considering how little result I see from AI, I wonder too.
Ads on the local airport use AI generated stuf, but that's the only example I've seen in my state.
Well, the answer is simple. Itās you sell it to addicts
People with problems whatever they may be social isolation heartbreak, loneliness, antisocial you name it
And not to clump everybody together Iām just saying if we look at the spread and weāre being honest with ourselves
ā²Bigtime addict, but the most I spent a month on AI clip generation is 10ā¬
For image generation, TTL, LLM etc = zero,
They say that thereās 2 billion AI users daily if not more
The only 3% pay for it
For YouTube, itās I think a little bit higher like three or 4% maybe five
So how do you extract the most of amount of money?
Oh yes, while nearly nothing seen locally.
The amount of AI slop on YT seem to be increasing badly.
Well, that 3% is holding up the AI industry for large language models at least
Think about that for a second
And the ironic thing is people that use AI consider themselves intelligent and a lot of them are
But if 98% of people donāt pay for it and youāre in that 3% lol
You think thatās the cream of the crop.?
I provided one example here the other day, with someone who apparently had used one of the larger AI's to do history research, used AI generated clips and AI voice.
The problem were the use of LLM to get the facts - which were 100% wrong, since the AI's use sources like wikipedia. š¹
You know, I never understood that expectation from anything especially chat box or large language model
I talk to people every day, dude that are so misinformed on so many things that donāt even use AI
So why would we think AI would be any different? lol
How many mistakes do you think the Internet has misinformation on websites news anywhere in general?
Google screwed up also, declaring their AI had found mathematical proofs that had eluded researchers for decades and longer.
Problem was that their AI accessed the single page of one researcher who had not updated for a decade - and the problems they thought their AI had solved were fetched from published papers. š¹
#video-arena-1 a lion dancing
You should see what itās doing to research papers and scientific papers
Hi guys iam also relatively new here and I want to create an ad with no camera no editing just pure ai what free resources you guys would you recommend to check out . š
The amount of people turning in fake research and discoveries has skyrocketed that thereās not enough people scientist and researchers, too verified all
is this true you can make unlimited video with this
No, Iām curious where you heard that though
who ever did not answer is gay
I am a researcher, I spot AI influence quite easily. If the author do edit, I might get uncertain.
Sweet thatās not me
Yes but in a limit only 5 vids per day
š
how long dose it take to make a video
š unnecessary but funny ...
and also thank you for answering
Lets go gay! šø
hahah
Around 5 mins but it only give you 8 sec of actually video
any body paly brawl star
Gay means happy - so no one should have any problem with that anyway.
But see thereās also a huge irony into everything I said
The thing is, I could be misinformed right now cause I donāt know
I mean, I donāt know no better than the next person
So everything Iām saying could be misinformation and I donāt even realize I didnāt think itās true or itās not
Indeed, and no worries.
It actually is a problem in research.
thank you khan saab you're a good guy do you know any good animator
As a viewer of my channel, youāll get an exclusive 50% off The Economist, but only until August 24th! Head over to https://www.economist.com/Sabine to get started.
Things are not looking good for the future of science publishing. Researchers are leveraging AI, paper mills are churning out more fake papers than ever, and people are constantly ...
What I meant to say is the information Iām getting is from the same source
The Internet, and then Iām criticizing it for having fake information
Perfect example of a non-scientist who mislead people.
So I just realized how the hell do I even know itās true and whatās not sure Iām just taking peopleās work for it that I see on YouTube or whatever itās no different
i need good image to video generator any body can help
Exactly, my opinion
Try grok itās free right now
and who ever did not answer is KSI
Brah thatās one way to get nobody youāve ever answer again š
Hossenfelder is one of the worst.
Debunked repeatedly by actual scientists.
I donāt know how some of you mf talk in real life lol but that would never fly
groak is free?
For now last time I checked, you could generate I donāt know how many number of videos 50 or something like that maybe Iām not sure
To generate it normally takes anywhere from 2-10 minutes to generate.
Nope I don't know
I was actually looking to make ads without any equipment and while surfing the internet I ended up here which is a good community everyone here is always ready to help
She's a nutcase.
And a whiner!
bro i cheak like 40 image generatro but there output is so poor
Well, whatās crazy is thatās who I got the information from
And this is what makes it really ironic
Yeah they nerfed it
This is why I canāt take anything serious from anybody or anything and thatās why my attitude is the way it is
Disinformation, misleading people with a false title she do not have, and utter lies is not 'free speech' - it's slander!
Cause Iām in the same boat as everybody else lol
Well, OK think about this so if sheās disinformation misleading people with false whatever
Then I wouldāve never knew about the fake science research stuff
Which makes it even that much more ironic that it ended up being true
What are you trying to generate?
I want to create an ad with no camera no editing just pure ai what free resources you guys would you recommend to check out . š Anyone??
Theyāre all the same kind of sort of take or minus
shorts
With a belt?
#1397655624103493813 has information on how to use the bot for video generation
wdm
It depends on how you give the prompt
You said shorts do you wear shorts with a belt?
That means you got a curious mind and have learned some stuf - good, keep it up.
Perhaps that might lead somewhere, lets say in AI research or even legal matters perhaps. šŗ
I donāt believe in intuitions or systems
why
I just think people are people
....
yes you are right
No matter where you go itās all the same in my opinion if itās China if itās North Korea if itās whatever
I just think itās all the same everywhere. I donāt know why I think people act the same everywhere.
ok you guys did not help but thank you gehlo for some help bye
Where are you going, bro?
Oh, come on, bro donāt be like that
wanna find ai
You need a map
ahahha
I know some shortcuts
There is a bot in this server that you can use to make video generations.
funny guy
Koi toh reply dedo ???
what bot
Just tell me what you need and Iāll help guide you in the right direction
Information is in #1397655624103493813
Start away here ali: https://discord.com/channels/1340554757349179412/1400148557427904664
im a short creator and i need to make imge and video a animal shorts
Could you help me out @keen beacon
I want to create an ad with no camera no editing just pure ai what free resources you guys would you recommend to check out .
you guys are not answering
cheak sora ai it is good
but you need invite code though
Ya
Best bet is to generate from an existing image then.
They try to make it exclusive, which is lame
Are you no longer approving requests for the Arena Champions Program?
It's been a week since I sent it
@echo aurora
ok bye
I have created some but it's very amateur .
Yes Iāll make it for u now
what wondering
We are, I tend to do those applications on Wednesdays
You get everybody stirred up and then you just dip
Bro have some patience ali
Then use that in another AI to clean up the image first.
I actually do that all the time - I do matte paintings to guide the image AI for the scene I want.
Didn't get you can you elaborate please
bro sora 2 is good
I have some animal videos
for ad is goat
Ya but it's paid though
I start with a simple image I draw myself.
This can be sent to an AI with is good to interpret line art images.
I now have a colour image, and add the text on top.
what model you use
make me a 3d anime of a muscular man in half-sleeved clothing with his fists outstretched as if ready to punch. with a background of an altar fight
Sora 2 is free I thought it' was paid will check it out then
for this video
but a speaker of truth?
yes cheak it bro but there is one problem with it
Nope. not in any sense of that word. She's just making things up to get attention on YT - period.
Oh I get it but how do you draw ?
Pencil and paper will do.
What?
how can you be so sure?
take an image with the phone camera and send to AI.
Iāll be right back Iām gonna make Connor video
Oh so you don't do it digitally
Go do your own research mate, there's numerous videos on YT by reputable scientists. And this subject is OT to the channnel anyway.
First sketch of mine is always with pencil - sometimes watercolour.
Oh what is that ?
who asking you i cheaked a lot of vido but people expermint is more valuebale
you need a inivite code to join there you can get it from open ai discord
Oh do you have an invite??
From where can we get the invite
No worries
bro you do not get what i am saying
i said from open ai discord
Sorry, sometimes my iPad is so full. I canāt even copy paste stuff.
So now Iām officially Iāll be right back. Iām gonna generate that for you guys.
gehlo give me the model name you make video with pls
I have send you a friend request will discuss it in dm
but i told you something this is my first time using discord
All right, what do you need? Iām back on you. Need funny animal videos or what
In the AI world there's many ways to turn simple lines into good images.
These are calles models / checkpoints - those two are the same thing.
Then there's loras that are used together with those above.
yes
All right, no problem
do you have sora?
i have sora 2
I sent you a request. Your request has been generated now.
but i do not have invite code
Damn, that sucks. I gave all mine away.
it is not good qulity i need
Khan wanted a vid
what you gave all mine away
sora gets the best quality
yes it does
Yep
you just have to prompt it right
no it's not
how you know if you havent used it lol
480 p is good kid
@keen beacon shall I dm you the script
And that's an art in itself also - especially as every AI got their own prompt interpreter. š¹
why are you insulting me when i was trying to help you, this is why i do not help people gonna ruin it for other people smh
dude is crying for help from someone else and then i ask if he has sora and he says he needs code so i send code and he starts insulting me lol, you cant make this crap up
oh okay
sorry if i make you said
We see the type every day, sadly - state of the net.
and you did not help me
you just saying sora makes a good quality
is that help to you
wow
Please be respectul when talking to other users. @earnest oasis
Hey everyone š
Iām building a training-generation program (education-focused content) running locally on a server with an RTX 4090 (24 GB VRAM).
Iām trying to find a good open-source model that:
fits within 24 GB VRAM,
performs well in French,
handles pedagogical or educational content generation,
and can process large context inputs (~64k tokens) from PDF files.
Right now, Iām struggling to choose the right setup ā between vLLM, Ollama, or other inference servers ā and to pick the best model for my use case.
If anyone has experience running large-context French models on a single 4090 or can suggest good pedagogical models / optimized configs, Iād really appreciate your advice.
š
Bonjour
Go talk to the Mistral guys mate!
bro w.e, imma know not to ever respond to you lol
hello 
bro sora 2 can only make 8 s video
what i can even do with that
and you want me to put that video with watermark
wow
So did you gave invite code of Sora in the chat
i have 15 second sora 2
even the watermark remover can not block it good
I give a sora 2 code invite for good answer
wow
yeah i removed it cause whats the point if dude is ungrateful lol
why u wanna remove it
api doesnt have watermark
I want it actually
bro it is anoyying
ill dm you
give the khan he need it
cant be putting it here anymore
Thank you
just go on reddit
but can someone tell about a good imge to video generator
they post Sora 2 invite codes
Theyāre all the same pretty much
Image-to-video leaderboards here - https://lmarena.ai/leaderboard/image-to-video
Compare models according to their ability to generate videos based on the given images.
u can do it with old famous people
crazy how sora is on this, i noticed its pretty bad with that
sora 2 actually sucks at image to video
no half the invite code in the reddit is used]yes it is
Ye
u need to see new posts
U can do anybody on grok
im trying to find a imge to video
Hmm what you mean? Sora doesn't have image-to-video on LMArena's Video Arena
in the app it does
Grok
No, only available through the Video Arena bot, which more info can be found in #1397655624103493813
guide me how can make them my self
ok
Put image add words wait
Yeah Sora does have the capability. However, with the bot it was running into issues so we couldn't add with the bot.
hope the output did not be poor
any short creator ?
here
can help me grow
haha sory for weird quastion
Dude, it doesnāt matter what tool you use it just how you use it, which is more


