Numan
144 posts

Numan
@RandomBriefs
Curious about everything Asking questions : to myself and to Grok 🤣 Just a regular person thinking out loud
Bergabung Mart 2026
94 Mengikuti17 Pengikut

This is very impressive. I have worked with teams across various countries and languages and been in situations where the other team spoke only one language, so it was very difficult to communicate our points. Even previous voice models were not good enough for active translation. But this seems to make it so easy and effortlessly possible. Working across different languages has become much easier and more comfortable.
English

Introducing GPT-Realtime-2 in the API: our most intelligent voice model yet, bringing GPT-5-class reasoning to voice agents.
Voice agents are now real-time collaborators that can listen, reason, and solve complex problems as conversations unfold.
Now available in the API alongside streaming models GPT-Realtime-Translate and GPT-Realtime-Whisper — a new set of audio capabilities for the next generation of voice interfaces.
English

@nikitabier This is very handy. And ask the posts on the ticker bundle together as well
English

drinking coffee and people watching in Europe is a top 10 activity
Roger Boylan@BoylanRoger
Morning coffee in Paris.
English

@immasiddx I can't seem to get my head around notion. its a bit complicated to use.
English

Using gestures on your iPhone to copy, paste, and control apps on your Mac is pure ecosystem magic.
My free app lets you switch between your favorite Mac apps with just a touch, minimize and maximize them with a swipe, and even use a grab-and-throw gesture to copy and paste.
choclift is now available with a big update on the App Store for iOS and macOS.
English

@Polymarket In last couple of days, I have almost everyday, seen a job cut news.
English

Will soon be flying to Europe and what I am worried about is EU Entry/Exit System (EES), saw crazy lines. Someone had uploaded a video in Instagram and it was probably the longest that I have ever seen anywhere.
But I am counting Gold Track to save me 🥲
What has been your experience, if you have flown out of an active EES airport?
English

This tool lets creators film freely and offload the heavy lifting of editing.
Key use cases:
- Raw footage analysis: Upload unedited clips—AI spots strongest hooks, dead zones, pacing issues, and visual highlights to suggest a tight edit roadmap.
- Competitor breakdown: Feed in viral videos to extract what makes them work (shots, timing, hooks) and adapt for your style.
- Script + structure help: AI maps visuals to narrative, suggests cuts, transitions, or B-roll to turn hours of film into polished content fast.
Filming lovers can create more without editing burnout.
English

You're spot on—most current tools lean hard on transcripts for efficiency. But this Algrow + Claude setup goes beyond that: it pulls key frames + visuals alongside the transcript, then analyzes hooks, dead zones, pacing, and why certain shots land (or flop). It's not native video "watching" like a true multimodal model yet, but it's a solid step up from plain text scraping. Real revolution comes when models process raw video streams natively.
English

@anitakirkovska @steipete The AI world is changing by the day.
Claude has strong capabilities, but Codex is generous with limits.
Loyalty in the AI world has no time to develop, tbh.
Especially when Claude reportedly tested removal of Claude code from the base plan.
English











