Creating the future with algorithms & data 📊 AI | Python | Deep Learning enthusiast Always curious, always building something new
Sacramento, CA
Joined March 2026
- Tweets2K
- Following7
- Followers28.5K
- Likes2.2K
Got the visuals. Now what about the edit? 🎬
Try Dreamina for $0.29 your first month, and get 60% off CapCut Pro or 40% off CapCut Ultra on Desktop from Oct 20–30. The offers apply to separate subscriptions with separate pricing.
Generate your assets in Dreamina first, then bring them into CapCut PC to finish your video with tighter cuts, smoother transitions, and polished captions and audio.
CapCut Black Friday: tinyurl.com/2v23rn22
Dreamina offer: tinyurl.com/5xbtv4x6
#CapCutDidThat #Dreamina #Smallbusinessmarketing #CapCutSale
More creative time for the price of one annual plan.
flova.ai/?refCode=3BFU8P57
Try separate conversations for different tasks with Flova 1.5.
flova.ai/?refCode=6RTSP48F
One project. More than one task.
One project, multiple conversations, more creative freedom!
- With Flova 1.5, run independent sessions for different creative goals
- Explore ideas in different directions with shared project context
- Switch conversations anytime and pick up where you left off
✨Turn ideas into videos, your way!
Arafat Rahman retweeted
One project. More than one task.
One project, multiple conversations, more creative freedom!
- With Flova 1.5, run independent sessions for different creative goals
- Explore ideas in different directions with shared project context
- Switch conversations anytime and pick up where you left off
✨Turn ideas into videos, your way!
🤖 Made with AI
Arafat Rahman retweeted
We’ve raised a $2.5M pre-seed to fix hardware hiring.
The most reliable way to know if an engineer can do the job is to watch them work.
In hardware, that’s been almost impossible to do at scale.
Colare runs simulation-based assessments that show how candidates actually work, in the tools they’d use on the job, before anyone books an interview.
We assess across:
- Mechanical (CAD, DFM, FEA)
- Electrical (PCB, schematics)
- Systems
- Manufacturing, & more.
Candidates solve problems from first principles.
Hiring teams see that across their whole pool, so engineers only interview people who’ve already cleared the bar.
We're already working with everyone from seed-stage startups to multi-billion-dollar aerospace manufacturers.
If you’re hiring hardware engineers, let’s talk:
colare.com/
Keep the creative momentum going with a workflow designed to handle conversations, assets, and tasks.
Flova 1.5. @Flovaai flova.ai/?refCode=MPLHELP7
Introducing Flova 1.5
Smarter. More Adaptive. Bring Ideas to Life.
✨ Run multiple sessions in parallel and explore your creativity freely
✨ Chat at any step and switch between creative tasks with ease
✨ Build and reuse Skills to keep expanding your creative capabilities
✨ Rediscover inspiration quickly with smart memory and retrieval
✨ Manage assets flexibly and keep your creative work organized
✨ Take control of your Agent and create at your own pace
🎊 Smart and flexible—turn ideas into videos, your way!
Arafat Rahman retweeted
Introducing Flova 1.5
Smarter. More Adaptive. Bring Ideas to Life.
✨ Run multiple sessions in parallel and explore your creativity freely
✨ Chat at any step and switch between creative tasks with ease
✨ Build and reuse Skills to keep expanding your creative capabilities
✨ Rediscover inspiration quickly with smart memory and retrieval
✨ Manage assets flexibly and keep your creative work organized
✨ Take control of your Agent and create at your own pace
🎊 Smart and flexible—turn ideas into videos, your way!
Arafat Rahman retweeted
a two-year-old mining company has raised $400M. @a16z, @khoslaventures and the CIA's venture fund are all on the cap table.
@jclozano9897 co-founded @MarianaMinerals with @tbc415 and @baker_tilney, two friends from stanford. it's now worth $1.6B. they buy mines after the market gives up on them, then run them with software and AI.
i asked him how you actually buy a mine, why AI's real bottleneck is metal, and whether he'd hire me for being good at claude code.
nontechnical, episode 3.
00:00 $400M and the CIA's venture fund
02:19 you can just buy a mine
04:41 2 years, $400M, $1.6B
08:58 cold emailing VCs vs. their first check
13:46 when to buy a mine (right after the market gives up)
19:54 every mining truck is a robot without a brain
25:08 everything human-made is grown or mined
26:15 the cheapest AI company wins
28:31 AI's real bottleneck is metal
30:55 america doesn't have enough miners
33:27 space mining and asteroids
43:13 how a $1.6B founder hires
46:17 would he hire me for claude code?
48:43 "slop cannon"
I’d love to explain an edit the same way I explain an idea: “Make these lines stand out, keep the shots flowing, and give everything the same feel.”
That’s what interests me about the CapCut Web plugin coming soon to ChatGPT. Starting with raw UGC footage, a simple creative brief could help guide the captions, transitions, and filters that turn it into a more polished product ad.
Early access is available on ChatGPT desktop, with mobile and web support expected in mid-November.
It’s a preview of the AI video workflows leading creators and OpenAI explored at CapCut World.
#CapCutDidThat #CapCutWeb #ChatGPT #CapCutWorld
🚨 Healthcare AI just got a new open-source contender.
Ling-3.0-flash-Sante is built for complex medical reasoning, deep research, and real-world workflows, not just answering basic health questions.
With 5.1B active parameters per token and a 90.47 BrowseComp score using multiple agents, this release gives developers plenty to explore.
Today, we’re officially open-sourcing Ling-3.0-flash-Sante.
Sante is a healthcare-enhanced MoE model with 124B total parameters and only 5.1B active parameters per token, delivering leading performance among similarly sized models on professional healthcare tasks and complex medical reasoning.
Beyond answering medical questions, Sante is built for real-world healthcare research and workflows, with strengthened capabilities in deep research, multi-agent collaboration, and tool use.
The model weights are now available — we welcome researchers, developers, and healthcare practitioners to explore Sante and build with it.
More below ↓
Eleven v4 Turbo leads four of the five new boards.
Eleven v4 leads the fifth, Bahasa, with Eleven v4 Turbo in second.
Introducing five new languages on the Voice Arena TTS leaderboard.
Models from @ElevenLabs rank #1 on all five of the new Voice Arena TTS leaderboards: French 🇫🇷, German 🇩🇪, Bahasa 🇮🇩, Tagalog 🇵🇭 and Indian English 🇮🇳.
Eleven v4 Turbo takes the top position on four of the five boards. On Bahasa, ElevenLabs holds both #1 and #2, with Eleven v4 first and Eleven v4 Turbo second.
Across five languages spanning Europe, Southeast Asia and South Asia, each judged only by native listeners, ElevenLabs finishes first on every board, ahead of the next-best model from another provider by 11 to 26 Elo points.
Voice Arena ranks models through blind, head-to-head votes by vetted native speakers, scored with Bradley-Terry Elo.
Board-by-board results below 🧵
Arafat Rahman retweeted
script based editing is live in HyperFrames Studio
edit long-form and scripted videos from the script itself
auto speech cleanup takes out the pauses and ums for you
then click any word and ask for a motion graphic right at that moment
get started ↓
Tom Tykwer working with Seedance on live-action footage is an interesting twist. He’s using real footage with actors rather than creating everything from scratch. And with the director of *Run Lola Run* involved, I’m curious to see how it turns out.
Arafat Rahman retweeted
What happens when a stage play meets AI filmmaking? 🎬
In BytePlus’s Next Frame, Run Lola Run director Tom Tykwer, director and playwright Bonn Park, and producer Sarika Hemi Lakhani explore Volcano High, their collaborative AI filmmaking experiment.
Classic data flywheel: collect failures, send to annotators, wait, retrain. OXR swaps annotators for a loop. It runs on docs with no ground truth, which a general VLM reviews for errors and shaky output; disagreeing rollouts go on the uncertain list; clustering traces shared causes; the VLM plans the next round and generates the fix data. First RSI in doc parsing.
Light independent layout model; recognition and ordering together. Leader on olmOCR-Bench, OmniDocBench and DocParsingBench. 101.0 s vs 683.1 s (Surya-OCR-2), single RTX PRO 6000. Layout replaceable.
github.com/SoMarkAI/OXR
Create the assets. Edit the story. Share your vision. 🎬
Dreamina's $0.29 first-month offer gives you a low-cost way to explore AI asset generation. CapCut PC's Black Friday sale makes video editing more affordable, with 60% off Pro and 40% off Ultra.
You can explore each offer separately, since Dreamina and CapCut have independent subscription plans and terms.
🔥 Oct 20–30, CapCut Desktop only.
CapCut Black Friday sale: tinyurl.com/2v23rn22
Dreamina + CapCut offer: tinyurl.com/5xbtv4x6
#CapCutDidThat #Dreamina #Smallbusinessmarketing #CapCutSale
Arafat Rahman retweeted
HeyGen Video 1 went up against today's leading video models in a blind test on @ArtificialAnlys.
Our first general-purpose video model ranked #1 in text to video.
Try it: platform.heygen.com/models/h…