Обожаю метро.

Оно добавляет мобильности и скорости, будучи в тандеме с публичным наземным транспортом и самокатами.

Метро запомнилось мне больше всего. Благодаря ему я спокойно передвигался по всему городу, колесил из одной точки к другой, успевая всюду и не переживая за время. В этом есть свой вайб, много людей и каждая станция особенная.

Я буду скучать по метро.

#thoughts #insights

AI with right tooling can literally help out while you're willing to offload some stuff.

https://youtube.com/shorts/tFBFzx5j94s?feature=share

If you remember, recently I've released a new Podcast Interview episode on YouTube which is more than hour long, but for social media I used to edit a shorter (reel/AD) video.

I found a solution where I combined Captions of specific timestamps and the ffmpeg tool to cut exact moments from the full video - without ever opening a video editor.

Here's how it works:
1. Take the transcript with timestamps
2. Feed it to AI - it reads through the entire episode and picks the most engaging moments
3. AI generates ffmpeg commands with precise start/end times
4. All clips render in parallel in ~2 minutes

Result: 18 clips + 2 trailer mixes from a 75-min podcast. In about 10 minutes. Zero manual scrubbing.

The interesting part? First batch was "informational" - facts, numbers, achievements. Decent, but forgettable. So I asked to find moments with emotion instead — humor, drama, conflict, personal stories. Night and day difference.s with emotion instead - humor, drama, conflict, personal stories. Night and day difference.

Turns out: for social media, emotion > information. AI understood that from text alone.

Full writeup on my Medium https://medium.com/@mecreate/how-i-turned-a-75-minute-podcast-into-18-instagram-reels-in-under-10-minutes-using-ai-cfdd3c319640

#AI #ContentCreation #ffmpeg #Productivity #VideoEditing #Podcast