Built a cheat code for B2B ads: AdClone AI
Remix any ad in LinkedIn’s ad library.
1️⃣ Upload your company website to scrape
2️⃣ Choose winning ad templates from Linkedin
3️⃣ GPT‑4o image gen delivers unlimited, on‑brand creatives in 30 seconds
We’re live on @ProductHunt →
https://t.co/IPgyqjeTzg 🚀
Built a cheat code for B2B ads: AdClone AI
Remix any ad in LinkedIn’s ad library.
1️⃣ Upload your company website to scrape
2️⃣ Choose winning ad templates from Linkedin
3️⃣ GPT‑4o image gen delivers unlimited, on‑brand creatives in 30 seconds
We’re live on @ProductHunt →
https://t.co/IPgyqjeTzg 🚀
@semrush in the images, it shows the llm being able to estimate a click through rate of the copy. do we have access to this LLM model? or where do we see the CTR in the app?
@domasbitvinskas yeah man, DM me i'm down to chat a little about it. i tried actually a bunch of different approaches (and approaches work/break depending on newest models) - but v clear rubrics and the "why's" really matter here. have to create frameworks to guide the LLM
@domasbitvinskas okay i got you. the answer is always the same when it comes to training models to be better than gpt-4...
create the best datasets in the world. create a lot of them. cover many cases. and prob pay people to do it (synthetic data sets aren't gonna cut it). it takes a lot of time