wow wow

I'm a huge AI enthusiast.

🚀 AI Product Founder & Tech Explorer
🔮 Building in the AI space
🤖 Sharing insights on AI products & implementation

[wow wow @wocaTvT]

@aiepeng @TabbitBrowser Thanks, but one issue remains.😮‍💨

[wow wow @wocaTvT]

@sekachov @atlasformac Atlas is great! Is there a shortcut to drag/pan a zoomed-in image? Looking for something like Figma’s Space + Drag. Cmd + Scroll works for zoom, but navigating the zoomed view is a bit tricky right now.🫡

[wow wow @wocaTvT]

I'll try🧐

[konstantinpaulus @konstipaulus]

We finally cracked it!

We built an open-source video editor that turns every edit into code.

I got tired of agents:
→ Losing context
→ Giving me outputs I couldn’t modify

So we flipped it:

→ Code is the new database.

[wow wow @wocaTvT]

@ego_agent thanks！

[wow wow @wocaTvT]

@ego_agent 😭It won't let me send you a DM either.

[wow wow @wocaTvT]

@BlumeDotCodes Which model in which agent is used for the analysis? I couldn't find this setting in the app.😶

[wow wow @wocaTvT]

😀🙌

[Jacky Kwok @ COLM 2026 @jackyk02]

Scaling self-verification with DeepSeek V4 Flash beats Claude Fable 5 on Terminal-Bench 2.1, while being 11x cheaper 💰

As open-source models become more capable, they can now generate large numbers of high-quality candidate solutions and verify their own outputs at very low cost.

For example, we find that sampling just 5 solutions with DeepSeek V4 Flash and ranking them using the same model with LLM-as-a-Verifier can lead to a significant boost in accuracy (79% → 88%), outperforming closed frontier models on Terminal-Bench.

Try it out today: https://t.co/UhudjKDaYI

More on verification scaling in my previous post.

[wow wow @wocaTvT]

@evpanferkin @thsottiaux Super unstable🫠

[wow wow @wocaTvT]

@atikcoder0 Agent👍

[wow wow @wocaTvT]

@viccpoes I'm eager to test - a paying user who started using Krea last year🥹

[wow wow @wocaTvT]

@paoloanzn CLIP😃😍

[4nzn @paoloanzn]

I just automated the entire "brain rot" video creation process that's making creators $50K-$100K per month on autopilot

can’t believe I’m about to give you the exact system for free…

these split-screen videos (podcast clip + Minecraft parkour) are absolutely DOMINATING tiktok and youtube shorts right now

faceless channels are pulling 6-figure months consistently

the psychological trigger is insane

the background gameplay creates unconscious dopamine hits that make viewers physically unable to scroll away

here's what my n8n automation does:

- takes any YouTube video/podcast
- analyzes transcript with AI to find the most viral moments
- auto-generates timestamped clips
- downloads and processes everything
- combines clips with random background gameplay
- adds auto-generated subtitles
- renders final videos ready for upload

the entire process that used to take creators 4-6 hours per video now happens in 5 minutes

ZERO manual work
ZERO video editing skills needed
ZERO content creation required

just feed it a podcast URL and you are done

the math is stupid simple:
3 videos per day = 90 videos per month
average 150K views per video = 13.5M monthly views

the workflow includes:

- AI content analysis agent
- automatic video downloading
- smart clip generation
- background video integration
- subtitle automation

this is the exact system printing money for the biggest faceless channels

and you're getting it for free

Comment “CLIP” + repost this + follow me and I'll DM you the complete JSON file plus the full setup guide

don't sleep on this

every day you wait is money left on the table

[wow wow @wocaTvT]

@damianplayer YT Agent😀

[Damian Player @damianplayer]

Here's a plug-and-play AI Agent that finds what’s actually trending on YouTube in any niche using n8n.

– Searches the last 2 days of content
– Pulls views, likes, comments, tags
– Filters short-form noise
– Stores & analyzes trends across 3 searches
– Returns links + a smart summary (not just raw vids)

Follow + Repost + Reply “YT Agent” and I’ll DM you the full workflow (free, no email bs).

[wow wow @wocaTvT]

@perplexity_ai 🤩

[wow wow @wocaTvT]

I'm on the waitlist for @perplexity_ai's new agentic browser, Comet: https://t.co/gjBInL9fzC

[wow wow @wocaTvT]

👍

[Joshua Xu @joshua_xu_]

We have incorporated HeyGen's avatar model to work seamlessly with Sora, and the results are truly next-level. This is probably the most advanced talking avatar video to date—outperforming real actors in many scenarios. With this, we can precisely fine-tune movements, expressions, and actions without the hassle of repeatedly reshooting real actors. Plus, there's no limit to how long these videos can be.

This isn't a digital clone—it's a fully AI-generated human avatar.

Check it out!