~Jay

Do-Be-Do-Be-Do

[Nihal Sarin @NihalSarin]

I would like to dedicate this win to @GmNaroditsky. It was one of his favourite tournaments. We have played endless bullets and hyperbullets, which definitely contributed to my growth as a player.

[~Jay @gjayanth]

https://t.co/bQiIyLQmJU

[~Jay @gjayanth]

A little crazy, a little weird and sounds fun (who doesn't like going to a spa) to get medical scans https://t.co/6A4AAsIFew

[NVIDIA AI @NVIDIAAI]

This #CVPR2026 paper from our research team is trending #1 on @HuggingFace 🤗

Meet LocateAnything: a vision-language detection model that rethinks bounding box prediction. For AI agents and robots, “seeing” is only useful if a model can pinpoint where something is fast enough to act.

Trained on 138M high-quality samples, LocateAnything decodes bounding boxes in parallel instead of one coordinate at a time, improving localization accuracy while dramatically increasing throughput for visual grounding and detection.

Project page: https://t.co/O7JMe8tzFM

[Mario Nawfal @MarioNawfal]

Mind-blowing moment from the DeepMind doc: Demis Hassabis gets told AlphaFold can predict all 1-2 billion known proteins in just a month.

He instantly says: “Do it.”

This is how history gets made

[~Jay @gjayanth]

Another wonderful piece. Love these long form articles.

[Jaya Gupta @JayaGup10]

https://t.co/5vDrndcmFn

[Anish Moonka @anishmoonka]

When a computer tracks the Indian classical dancer in this video, it picks up perfect circles, triangles, and curves in every movement. There are exactly 108 of them. All 108 were written into a manual over 2,000 years ago.

That manual is the Natya Shastra. Six thousand verses, written somewhere around 200 BCE. It describes 108 specific dance movements for Bharatanatyam, one of the oldest dance forms in India. Each movement spells out three things: where your hands go, what angle your body holds, and the exact path your legs trace. Roughly 150 step combinations grow out of those 108 base movements. A trained dancer spends years learning 70 to 80 of them.

Watch the dancer's legs in the video. The bent-knee squat creates a diamond shape. Palms together make a triangle. When researchers plotted these positions in three dimensions this year, they found the moving body carves out twisted spirals and bowl-shaped curves, the kind of shapes you see in an engineering textbook, not a dance studio. Every limb holds a specific angle and moves a measured distance.

The rhythm is math too. A 7-beat song gets filled with dance steps of 3 and 4. Scale that to 35 beats and the groups of 3 and 4 repeat five times. Choreographers work out these splits in their heads while performing live. All 108 movements are also carved into the stone walls of a 12th-century temple in Tamil Nadu called Chidambaram, many panels still carrying the original Sanskrit description next to them. A choreography textbook in granite, still legible after 900 years.

A 2013 study put 25 people on a walkway rigged with motion-capture cameras. Every human stride has two parts: when your foot is on the ground and when it swings forward. The ratio between those two parts came out to 1.620. The golden ratio is 1.618. Your foot lifts off at 61.8% of every step you take, and it has done this your entire life. A Bharatanatyam dancer takes that same built-in proportion and amplifies it across 108 movements, each one tracing shapes that were set down in writing over 2,000 years before the tracking software in this video existed.

[~Jay @gjayanth]

One of the best articles I have read on the fast changing workplaces in the world of AI. Mandatory read for young (and experienced) people.

[Jaya Gupta @JayaGup10]

https://t.co/yiGW8jKKXk

[~Jay @gjayanth]

Sir - Your text books on Linear, Non-Linear Control Systems and Robotics were like Gita during my Masters. You are legendary and an inspiration to a whole generation of scientists. 🙏

[M_Vidyasagar @stellensatz]

I had been following and enjoying Shri Parimalji's articles: Remembering some old friends, and discovering some new people. I never dreamt that *I* would be the subject of one! Many thanks for the kind consideration. 1/n

[Anil Kakodkar @anil_kakodkar]

Today is a historic day. India has entered 2nd stage of our three stage nuclear power program with the achievement of clriticality of PFBR. Congratulations to every contributor to this critical technology that makes India only the second country to operate a large fast reactor.

[~Jay @gjayanth]

Is this for real? Crazy indeed.

[Dr. Datta M.D. (Radiology) ✈️ Switzerland @AI4Good @DrDatta_AIIMS]

How do they produce such crazy papers?

[Arun 🌞 @arunv2808]

This is actually a big surprise!

Great going Sarvam team.

https://t.co/wphm9v9aJk

[Nikhil Hooda @_nikhilhooda]

i’ve published to neurIPS but i still can’t understand research papers

that’s why we made arXivisual to make every paper feel like a @3blue1brown video

🏆 winner @ TartanHacks ‘26

link: https://t.co/4noZCuhxTy
repo: https://t.co/Zq4ngYCVYB

built with: @AjithBondili @_rajshah6 @armaangupt0

[Abhishek Maiti @o_v_shake]

People at India AI summit: if you want to work at frontier labs and dont want to spend money to prep for interviews, here’s a new resource: https://t.co/vKyfr2vIlN

[sierra🇮🇳 @sierraoperator]

the aesthetics of republic day parade 2026 1/n

[~Jay @gjayanth]

I’m at the AWS AI Conclave Bengaluru

Let’s catch up if you’re around.
#awsaiconclavebengaluru

[~Jay @gjayanth]

@calcsam BOOK

[Anders Sandberg @anderssandberg]

OK, color me officially impressed: Nano Banana Pro can make good diagrams based on papers. This one can go straight into my presentations.

[Deedy @deedydas]

If you feel like giving up, you must read this never-before-shared story of the creator of PyTorch and ex-VP at Meta, Soumith Chintala.

> from hyderabad public school, but bad at math
> goes to a "tier 2" college in India, VIT in Vellore
> rejected from all 12 universities for US masters despite 1420 on the GRE
> fuckit.jpg
> goes to the US anyway on a J-1 visa to CMU with no plan
> applies for masters (again) to 15 universities
> rejected from all except USC and with late admissions, NYU in 2010
> finds this guy called Yann LeCun (before he was famous)
> starts getting into open source
> rejected from all jobs including DeepMind
> only job is Amazon as test engineer
> his PhD mentor helps him get a job at a small startup (MuseAmi)
> rejected from DeepMind
> couldn't get H-1B because of J-1 home return issue; gets waiver through months of approval with USCIS and US State Dept
> very low on confidence
> In 2011/12 builds one of the fastest AI inference engines on phones
> rejected from DeepMind
> emailed Yann again and joins FAIR because of Torch7 open-source work
> scrapes through bootcamp at Facebook, struggling on an HBase task
> L8/L9 engineers at Facebook struggle to get ImageNet working
> figures out numerics / hyperparam issue as an L4
> first big win!
> FAIR goes well, runs 3 person torch7 team and co-creates PyTorch
> because of politics, management wants to shut down PyTorch
> cries-at-bar.jpg, literally
> eventually some people save PyTorch and it launches in 2017
> gets a EB-1 green card!
> the rest is history...

Think about that. He went to a tier 2 college. Was rejected from all Masters programs 2x. Rejected from every single job except Amazon test engineering. Rejected from DeepMind 3x. Nearly had his baby project shut down. Struggled with visa issues. After 12 years of failures (2005-17), he eventually rose to became a VP at Meta one of the most influential people in AI!

Soumith's story is one of resilience and he's living proof that no matter how down in the dumps you are, there's always hope.