
building and pushing products to market | puzzling inside @clipthisapp
Joined June 2015
- Tweets201
- Following571
- Followers237
- Likes7.9K
you can hate kevin o leary or love him whatever
but if you're the guy whos debating him you're never gonna make it.
- victim mentality
- keeps interrupting and cant listen to people probably smarter or in a better place than him (ego problem)
he applied on 100 jobs and got only 4 interviews and was complaining
I see that as a 4% interview rate which is now just a volume/skill issue. Either do 1000 applications and get 40 interviews, or increase your skills to increase your interview %rate
people want security guys that are AGILE , STRONG and INTIMIDATING. In that workline your appearance and body are your resume.
If you dont like that just change your workline.
teapaw.org/KillaKreww/status/2122…
Kevin O’Leary gets CHALLENGED by a security guard who works 80 HOURS A WEEK after arguing that ANYONE can become WEALTHY through hard work, only for the guard to reveal that 55% of his income goes to RENT and he’s applied to HUNDREDS of jobs without success, questioning how much HARDER he’s supposed to work 😲👀
12
368
It has become a routine now.
I’ll not even call it AI UGC anymore, just the new UGC format*
this shows that the 1million video was not LUCK, as we’ve seen 2 other videos right after it have very good results. (330k and 100k views) all via our same process.
high retention and engagement videos
Idea > Generation > Post Editing > Posting
it sounds simple on paper, but the process I’ve built behind @ClipThisApp to generate these flawless videos is art.
if you wanna start running creators for your brand, DM me.
HOLY SH*T my AI UGC just did 1M views in 22 hours... and it's still going
here's exactly how I did it with @ClipThisApp team:
> find a format that already has views. we rebuilt a 7.6M views tiktok: a girlfriend filming her bf staring at his trading screen
> wait for a moment your audience is living.
the market crashed 5th November after the election results. idea at 8:45PM that night, posted within 2 hours.
> our backend system that we use for generating real looking UGC (generating, editing, audio processing, using 100s of viral content as reference and video look. etc...)
> the more boring it looks, the more real it feels
> keep it short enough to loop, 11 seconds. people watched it 1.6 times each. as they were looking at the comment sections, video was looping 2 times in average.
why it hit:
we’d posted this exact format before, on a random day. it got 2.8k views
same format, posted hours after a crash = 1M.
people sent it to the friend who lost money that day, 6.7k shares vs 5.9k likes. the comments turned into everyone posting their own losses, or being the one to criticize our character for losing money.
CRITICISM + RESONANCE
the girlfriend telling the story makes it a skit, not an ad. barely any of the 437 called it AI
this is what we do at clipthis studio...
if you have an app & want this, reply “UGC” & i’ll DM you
HOLY SH*T my AI UGC just did 1M views in 22 hours... and it's still going
here's exactly how I did it with @ClipThisApp team:
> find a format that already has views. we rebuilt a 7.6M views tiktok: a girlfriend filming her bf staring at his trading screen
> wait for a moment your audience is living.
the market crashed 5th November after the election results. idea at 8:45PM that night, posted within 2 hours.
> our backend system that we use for generating real looking UGC (generating, editing, audio processing, using 100s of viral content as reference and video look. etc...)
> the more boring it looks, the more real it feels
> keep it short enough to loop, 11 seconds. people watched it 1.6 times each. as they were looking at the comment sections, video was looping 2 times in average.
why it hit:
we’d posted this exact format before, on a random day. it got 2.8k views
same format, posted hours after a crash = 1M.
people sent it to the friend who lost money that day, 6.7k shares vs 5.9k likes. the comments turned into everyone posting their own losses, or being the one to criticize our character for losing money.
CRITICISM + RESONANCE
the girlfriend telling the story makes it a skit, not an ad. barely any of the 437 called it AI
this is what we do at clipthis studio...
if you have an app & want this, reply “UGC” & i’ll DM you
Riad retweeted
agorian.ai is live.
Get the market’s lowest-cost inference. Fast, reliable, and scalable.
Designed for you to run more models, serve more users, and keep quality high while spending less.
Full rollout on Monday.
its getting tiresome of people just farming X engagement for vague posting or whatever
can you do this sort of UGC quality with AI?
100% yes and thats what I’ve been doing in @ClipThisApp
but if you think this can be done via a simple prompt and casually get this quality, you are in for a suprise.
> Character creation
you can use nano banana or gpt 2.5 to create the initial look, but those will never look good in seedance or omni or wan. First you really have to nail the look in detail and create multiple short VIDEO scenes as thats what makes the character look good and to not look like its recorded from a professional camera or have that focus blur that most generations have.
Then you have a starting point from that.
> Audio
Sure, for ads maybe audio does not matter, but in UGC organic content…it does.
There is a multi-process workflow to edit the audio so voice does not have that Echo feel Seedance and all other models give.
Then lip sync. Seedance is best at it for the moment.
> Emotion
If you use audio from Elevenlabs or any other tool, it sounds clean but flat. Real people dont talk like that. They pause in weird places, breathe, restart a sentence, look away when they think.
Thats why in our process we have real people record the actual UGC first. The voice, the pauses, how the phone is held, where they look. So the model knows from the start what it has to look and sound like, and then we turn that into the character.
The voice has to carry the emotion first and the video has to follow it, not the other way around. And you have to tell the model what he does between the words, a small hand move, a nod, a look off camera. Otherwise you get that dead stare into the lens that screams AI.
> Camera
The biggest giveaway is that AI videos look too good.
Real UGC is shot on a phone. Someone holds it, it moves a bit, the exposure jumps, the framing is a bit off, the room light is not perfect.
So every shot has to be written like someone actually filmed it. Who is holding the phone, where are they standing, front or back camera, what light is in the room.
> Script
If it sounds like an ad, nobody cares how real it looks.
Write it like a voice note to a friend. Short sentences, real numbers, no slogans. The first 3 seconds have to tell you what it is and why you should stay.
> Edit
Generation is maybe half of it.
Cutting on the right words, voice matching the lips frame by frame, real app screens instead of fake AI ones, captions in the native font, zooms that actually mean something. Thats where it starts to feel like a real creator and not a demo.
> Checking
Every clip gets checked before it goes out. Hands, eyes, logos on clothes, text on screens, lip sync, echo in the voice.
One thing off and the whole video feels off.
So yes, it can be done. But its not one prompt, its a full pipeline and a hundred small details nobody sees.
thats what we are building at @ClipThisApp
this is why I love AI right now
back in the day all goated tools were essentially mostly wrappers, but noone was gonna spend the time to reverse engineers their workflows or system
but those days are over. and seems like companies are not happy about it
it’s truly PVP out here now
Last night, @higgsfield sent @ComfyUI a cease and desist (attached)
They want the Open Source Higgsfield challenge taken down within 48 hours.
Their argument is that our use of the Higgsfield name and that using their CEO's likeness is an infringement violation (man, the irony). Claiming that it makes the event look like Higgsfield is associated with it.
This couldn't be farther from the truth, which is that Higgsfield made a promise to open source its product and never did.
For the past two weeks, builders have been recreating popular closed apps as open workflows anyone can run.
So we asked our community to deliver where Higgsfield failed: apps in the open, built by the community, for the community. We've made it clear that the challenge is organized by ComfyUI and not affiliated with or endorsed by Higgsfield.
When we started Comfy, we made the community a promise. Our product stays open, free, and yours. We’ve kept it every day since, and one letter isn’t going to change that.
The challenge stays up. Comfy and open source will win!
3
70
man... this is not going to be used for serious reasons
men are gonna get into DEBT to get this in their homes
teapaw.org/RealbotixCorp/status/2…
The wait is over. Meet #Echo humanoids.
Introducing Realbotix's first commercial robot line, available in limited quantities.
Choose from four distinct characters, each with its own face, voice, and personality. Available in tabletop and full-body configurations, Echo offers a range of options to choose from.
Say hello to Mila, Aria, Dave, and Amar.
Discover Echo: realbotix.com/
🇨🇦 #TSXV: #XBOT $XBOT.V 🇺🇸 #OTCQB: #XBOTF
#Realbotix #HumanoidRobots #EmbodiedAI #SocialAI #PhysicalAI
don't wanna be a party killer, put there is a open source repo right now for quite a while that has reversed engineered Genjutsu from Higgsfield which all of you can use yourself
all you need is a Seedance api key and you can create your own custom workflows with it pretty sick and cheaper. github.com/sirioberati/Genju…
use claude to figure it out
teapaw.org/davidim/status/2114764…
We just killed Higgsfield Genjutsu.
Introducing Sume Remix : same workflow as Genjutsu, 2.6x cheaper
> clone any video
> no markups, no subscriptions
> start from $5
Try now at sume.com/remix
alex is so right here
China is already ripping with AI dramas and it's not even close... study Dreamwave
if you have good storytelling skills, and know how to direct AI filmmaking sky is the limit
dramas, full films, ads, skits, education
teapaw.org/ZssBecker/status/21136…
FYI most php2shell exploits/POCs you will see in github are either backdoors/honeypots or malware.
so no... you're not on your hacking era trying to be a script kiddie right now, be careful
1
8
141
instead of seedance now lite version uses wan3.0 model I assume.
This is very easy to replicate it yourself as well...
teapaw.org/higgsfield/status/2112…
Introducing Higgsfield Genjutsu Lite.
Drop your photo into any trending video and you're in it.
Generations cost 3.5x less at 720p and 5x less at 1080p.
Available now on Higgsfield and via API. Genjutsu and Genjutsu Lite are free to try in select regions.
2
90
elysium is basically Base for hyperliquid.
BUT most eth L2s pull value away from the main chain
this one is built to bring activity back into hyperliquid
it will run on $HYPE and targets 100x the capacity of hyperevm
eyes are open👀
teapaw.org/Enter_Elysium/status/2…
7
125
sad to see what happened to $PAPER but understandable from the team.
the manipulation that was possible from perps and then token enabling to transferable makes the decision make sense.
however, when something seems to good to be true and everyone is in usually means one thing....
congrats who farmed early staking rewards tho
3
80
the only thing left to make AI gen content indistinguishable from real is the audio.
once that is nailed it's over
that's why right now I still prefer making performative AI content where I can replace it's audio to have a more real sound.
but tbf in most cases the usual user doesnt notice it much
teapaw.org/minchoi/status/2111471…
Just 48 hours ago Kling 4.0 opened to everyone.
These are not real🤯
7 wild examples + PROMPTS:
1. UGC character at brunch
PROMPT: Raw vertical 9:16 video recorded on a phone's front camera (not on screen), one continuous 30-second take with no cuts. It looks like an ordinary unedited phone clip, not a film: slightly soft image, a little digital noise and compression, deep focus with the café behind her as sharp as her face, auto-exposure brightening and dimming as people pass the sunny window behind her, no colour grading. The phone sits close and a little low, looking slightly up at her; around 9 s a waiter's arm bumps the table, the phone slides and tilts, and she catches it and props it back up. Sadie, a young woman in her early 20s with long honey-blonde waves, light natural makeup, small gold hoop earrings and a cropped cream knit top, sits at a small marble table by a big sunny window in a busy brunch café. Her sister sits across from her, out of frame except for her arm and a hand with a silver ring. Behind Sadie: other tables of people talking, a counter with a glass pastry case, plants, staff moving about, chalkboards too far away to read. Around 9 s a waiter sets down two plates, a lemon ricotta pancakes and a plate of eggs on toast, plus two iced coffees in plain glasses. Around 15 s she turns the pancake plate toward the camera; around 20 s she cuts a bite, eats it and closes her eyes; around 26 s her sister's hand reaches in with a fork and steals a piece of pancake, and Sadie swats at it, laughing. [Sadie: casual young woman's voice, warm and a bit giggly, recorded on the phone's built-in mic from half a metre away]: "Okay, I promised I'd show you what we get. ... We get the same thing every single time. ... Oh, here it is! Thank you! ... Look at this. Lemon ricotta pancakes. ... Mmm. Okay. Yeah. Still the best. ... Hey! Get your own!" [Her sister, off camera: a young woman's voice, dry and amused]: "That's the tax." [Sadie, laughing]: "She does this every time." Audio as captured by the phone mic in a busy café, not studio clean: her voice close but with the room around it, the hiss and grind of an espresso machine, cups and cutlery clinking, overlapping chatter and laughter from other tables, a door bell jingling, a chair scraping, faint music in the background. No subtitles, no captions, no on-screen text.
7
57
2 scenarios I see for the future of consumer PCs
1. prices eventually go back to normal
either AI gets more efficient, chip production catches up, or demand drops down enough that regular consumers can afford hardware again
2. we just stop buying PCs altogether
you buy a monitor with a basic chip inside, pay a monthly subscription, and stream your entire PC from the cloud
I think we’ll eventually see both, but my hope is option 1 does have a chance of happening
Just found out a RTX 5090 now costs $12,000 USD. I’m waiting to build a 10950X3D2 + 6090 PC next year, I’m guessing the entire build will be 25-30K with the way things are going.
Question, what happens to the average income consumer? Something has to give eventually right?
4
66



























