Jerry Liu retweeted
Replying to @tfmlab
Thanks! Curious how you're building those short AI explainers. Are you using Opus 5.5 or Fable 5.2? Or something like Higgsfield, OpenArt, or Fal via API?
🚨 BREAKING: The movement to bring Jensen Huang back as NVIDIA ($NVDA) CEO has crossed 100,000 signatures!
The numbers are remarkable:
53,000+ signers identify themselves as NVIDIA shareholders.
6,200+ identify themselves as current or former NVIDIA employees.
@elonmusk is among the lead signatories.
Fundstrat’s Tom Lee and Shopify CEO @tobi have also signed.
The message is clear: Jensen’s leadership and 33-year legacy matter to shareholders, employees, and the broader AI community.
If you stand with Jensen, add your voice here:
✍️ Sign the petition: standwithjensen.com/
#StandWithJensen #NVIDIA #NVDA #AI
I make Transformers and AI fun and easy to understand, with 3D animations, demos, real use cases and AI news.
Every morning: one animated Evolution Grid, six upgrades side by side and when to use each. This one shows how attention lets every field of a card payment look at every other field.
If you build LLMs, large transaction models, or fraud or credit models, say hi. I reply to everyone. #MachineLearning #LLM
THE DISHWASHER WHO CAME BACK TO DENNY'S
Jensen Huang (@JensenHuang), Part 2 of 5
At fifteen he washed dishes at a Denny's, too shy to talk to customers.
Fifteen years later he sat down in a Denny's booth again. What was planned in that booth changed computing. ↓
"I was incredibly shy."
"I was horrified by the prospect of having to talk to people."
Dishwasher. Then busboy. Then waiter.
Then, in a booth in San Jose with Chris Malachowsky and Curtis Priem, they planned @nvidia.
"To me, no task is beneath me, because, remember, I used to be a dishwasher."
We stand by @JensenHuang.
The plan: one part every evening at 6 PM ET.
Fri · 2. The dishwasher's booth ← tonight
Sat · 3. Thirty days
Sun · 4. The bet nobody wanted
Mon · 5. What the pain was for
(Part 1, The tape, is already out.)
If you build with AI, or his story means something to you, let's connect.
Watch Part 2 (2 min):
youtube.com/shorts/HQNVJO7Jf…
Jensen Huang and his brother spent about two years talking to their parents through one cassette tape.
Every month they sat in front of a tape deck and told their mom and dad what they had done. Their parents recorded over the tape, and mailed it back.
1973. He was nine, by his own telling. There was unrest in Bangkok. His parents sent the two boys across the world, to a school in Kentucky, to keep them safe. They stayed behind.
His job at the school was the bathrooms. For a hundred boys.
"It wasn't a punishment. It was just my job."
His roommate was seventeen and could not read. Jensen taught him.
Fifty years on, he remembers that school as good. Then he gets to the tape:
"Probably the only thing that really breaks my heart."
And this is what he says he is:
"I'm the product of my parents' dreams and aspirations."
Before the chips, before @nvidia , there was a boy and a tape.
We stand by @JensenHuang .
His story in his own words, in 5 parts, one a day:
1. The tape
2. The dishwasher's booth
3. Thirty days
4. The bet and the delivery
5. What the pain was for
Much of it he told to @joerogan .
Part 1 (2 min) ↓
youtu.be/h14s5YYjA8s
Jerry Liu retweeted
🇺🇸🇨🇳 Hillary Clinton just raised a VERY uncomfortable question about the Nvidia spy scandal...
If U.S. intelligence suspected Jensen Huang was being targeted in a Chinese “honeypot” operation, why wasn't he warned?
Clinton argues the government had a duty to tell him.
Instead, she claims officials allowed the relationship to continue while encouraging Nvidia chip sales.
Writer: Samuel
Jerry Liu retweeted
Gemini Robotics 2.5 is here. 🤖
Our latest physical AI is able to learn hundreds of everyday skills - from folding laundry and serving pastries to autonomously sorting car parts on the factory floor.
Here’s a snapshot of its capabilities 🧵
Happy Sunday ☀️ Let's connect: AI, ML and LLM engineers, solo builders, AI leaders.
Today:
- ML interview Shorts: LR schedules, cross-validation, missing data
- breaking AI news + our take
16-min Sunday watch: the Transformer, attention to DeepSeek
youtu.be/pXlH5depzsw
Replying to @JerryWangKK
Great pick, and we've animated exactly that: one new token arrives, matches the cached keys, reads the cached values, and the earlier ones never get recomputed. Our workflow is code-driven too: every frame is computed in the browser from the real maths, then voiced and rendered. Will read your breakdown, Jerry.
Thank you, fans! 🙏 From 130 to 578 followers in 3 days.
Great work gets rewarded, and it began with a 16-year-old reading Transformer papers.
We'll keep building the best visual AI knowledge so you learn fast, with fun.
Let's connect, build and have fun together, one spare minute at a time.
My 16-year-old daughter now reads Transformer papers for fun. Not for school. The animations made "Attention Is All You Need" click for her: every word looks at every other word, all at once.
Reading papers can be fun. If you work in AI, learn the Transformer first. It is the engine behind almost everything now.
Here is the 2017 paper and what came after it, in one short film. Learn AI in your spare minutes. Let's connect here.
BREAKING: Hillary Clinton on Elon Musk: "Let him pardon Elon and then let's call him in." Her point: a pardon can't stop a congressional subpoena, and once pardoned you can't plead the Fifth.With Democrats now running both chambers, Musk is in their sights.What does it mean for AI leaders and the people building the future: protect them, or hold them to account?#AI #Tech @elonmusk, what do you think?
For everyone new to this channel: #Transformers are at the heart of today’s AI revolution!
This is arguably one of the most important research papers in modern AI history — Attention Is All You Need.
I turned this groundbreaking paper into an easy-to-follow 3D animation so everyone can understand how Transformers work, without getting lost in complex mathematics.
Follow this channel, there will be more groundbreaking animation for you to learn
She liked the first animation, so we went one level deeper for her: the whole paper in 3D.
Words become points in space, attention draws the lines between them, and the encoder and decoder stack up layer by layer. Same paper, Attention Is All You Need, now something you can watch.
AI learners, AI leaders, business builders: connect, learn something new in a spare minute, and enjoy it. We grow faster when we learn together.
#AI #MachineLearning #Transformers #LearnAI
Want the full course? This 16-minute film walks through what came after this paper, era by era: attention, BERT and GPT, RLHF, RoPE, KV cache, mixture of experts, all the way to ChatGPT and DeepSeek. youtu.be/pXlH5depzsw
ML interview classic: XGBoost vs LightGBM vs CatBoost. Same gradient boosting, different trees: level by level, the leaf that cuts the error most first, or the same split across each level.
Preparing for ML interviews? Learn it in your fragmented time. Let's #connect and grow together.
I never thought I would have an X with 170k impressions on day5! It means you do love to learn #AI visually and know AI news interactively with insights.
Thank you!
Let’s connect!
I promise it is going to be a rewarding experience using your fragmented time!
#connect #AI @tfmlab
Let’s connect, learn together and have fun!
My 16-year-old daughter now reads Transformer papers for fun. Not for school. The animations made "Attention Is All You Need" click for her: every word looks at every other word, all at once.
Reading papers can be fun. If you work in AI, learn the Transformer first. It is the engine behind almost everything now.
Here is the 2017 paper and what came after it, in one short film. Learn AI in your spare minutes. Let's connect here.
Replying to @Revolut
@Revolut's AI-ops lead says moving 2,000 engineers to a self-hosted open model cut the AI coding bill to about a tenth, and the surprise was where the time went: coding wasn't the bottleneck, rationed tokens were.
Would your team trade the newest model on every task for unlimited tokens?
My 16-year-old daughter now reads Transformer papers for fun. Not for school. The animations made "Attention Is All You Need" click for her: every word looks at every other word, all at once.
Reading papers can be fun. If you work in AI, learn the Transformer first. It is the engine behind almost everything now.
Here is the 2017 paper and what came after it, in one short film. Learn AI in your spare minutes. Let's connect here.





