Video yükleniyor...
Video Yüklenemedi
like okay? 😭
37 Yorum

update: ????????????????????????

If you analyze the hash of all media you can then see the ones that are the most re-used and sort by longest running to see the most likely best performing ads (Or you can just use AdKit to do it for you and also launch everything on meta later 😀)

good point but meta ads library says how long it's been running for no? or is that something else

You can calculate it, starting date VS ending date!

generational print

i'm about to make 500 creatives

bro I might have to try this, crazy

the levels of cookery is insane, from my latest batch

crazyyy

yeah it's going to be a long day today lol

fr fr

What about the ss of the app and the animations? Did you supply them/ did claude access the app and get em/ were they hallucinated?

used my simulator to record it bro

Makes sense

Do u actually write prompts like that?

depends on my workload, i have 8 sessions running across web, backend and frontend so i just send it, why

The “up for a challenge” parts seem pretty tedious if your writing every prompt in a casual style like that. Just seems faster to remove those and get the same output.

nah i dont prompt like that i'm just sleep deprived and excited bro this shit slaps, also i'm a regard

😭love it bro

🔥🔥🔥

lfg

insane

agree it went to hard

What yacht are you going to buy?

the one that can fit my porsche

That’s what I like to hear.

Man this means no sleep and ads 🔥🔥🔥

fire up the machines brother

Did it record the app in the simulator, or recreate the UI for the ad?

record in the sim!

so opus 5.5 is your creative agent it’s probably tripping out but making hella dope stuff ngl

sick. how does it watch/understand videos?

basically sent out 10 agents scraping all competitors videos on meta ads library and analyzed the videos frame for frame, here is an image from a command it ran then dispatched other agents to analyze hooks etc, crazy stuff

holy sheet

yeah just woke up time to kickstart the factory

Dope

insane
