Ajit Tripathi

380 posts

Ajit Tripathi

Ajit Tripathi

@triptananda

AI/Machine Learning and all things science, mathematics and deep tech enthusiast alt of @chainyoda

San Francisco, CA شامل ہوئے Mayıs 2010
557 فالونگ60 فالوورز
Ajit Tripathi ری ٹویٹ کیا
Marc Andreessen 🇺🇸
Silly Business Theory is right: in the future the best work, greatest progress, most valuable innovations won't come from laboring under the false consciousness that work must be hard and serious to produce value. The best work is going to come from people playing and having fun.
47fucb4r8curb4fc8f8r4bfic8r@47fucb4r8c69323

The more I look at this the more impressed I am and the more I realize how grateful we should be to Tao. 1. He acknowledges ignorance: this is something academics almost never do since their cultural capital is tied up in them knowing things. But he can since, well, he's Terence Tao. 2. He is explicitly acknowledging his use of GenAI to fight the stigma of using AI. If the child prodigy turned UCLA prof who studied with Erdos uses AI, it is legitimate technology. (please start using this sentence with AI skeptics btw) 3. He is also showing how AI is best used: as a kind of syntactic tool that finds connections in possibility space and has access to a larger library of information than our brains can. There's more here but the cool internet thing is a list of three. I often lament Tao has too playful of a mode of operating, feeling like he plays with linear algebra when he should be doing foundations of mathematics. But not only does this moment prove my view wrong, it also proves just how much Silly Business Theory #SBT is right: in the future the best work, the greatest progress, and the most valuable innovations won't come from people laboring under the false consciousness of Protestantism and Marxism that asserts work must be hard and serious to produce value. The best work is going to come from people playing and having fun. We're on the cusp of a near utopian explosion in human potential and quality of life. And you're bearish?!?!?!?!??????!?

English
125
203
2.5K
413.3K
Ajit Tripathi ری ٹویٹ کیا
chainyoda
chainyoda@chainyoda·
Things working well for young people in the age of AI - building cool side projects they are interested in - publishing papers without paying the university system - selling and pitching - networking - retardmaxing - posting demos, opinions, ideas and news on x and insta
English
5
3
41
1.4K
Ajit Tripathi ری ٹویٹ کیا
Neil Tripathi
Neil Tripathi@tripathi_neil·
This is so cool
English
13
4
30
2.4K
Ajit Tripathi ری ٹویٹ کیا
Andrej Karpathy
Andrej Karpathy@karpathy·
- Drafted a blog post - Used an LLM to meticulously improve the argument over 4 hours. - Wow, feeling great, it’s so convincing! - Fun idea let’s ask it to argue the opposite. - LLM demolishes the entire argument and convinces me that the opposite is in fact true. - lol The LLMs may elicit an opinion when asked but are extremely competent in arguing almost any direction. This is actually super useful as a tool for forming your own opinions, just make sure to ask different directions and be careful with the sycophancy.
English
1.7K
2.4K
31.2K
3.4M
Ajit Tripathi ری ٹویٹ کیا
Neil Tripathi
Neil Tripathi@tripathi_neil·
Just built my first SO-ARM101 robot arm! LeRobot + HuggingFace made it really straightforward, reminded me of how much I loved my Lego days. Leader arm next, then training imitation learning policies 🤖
Neil Tripathi tweet media
English
1
2
12
641
Ajit Tripathi ری ٹویٹ کیا
Beff (e/acc)
Beff (e/acc)@beffjezos·
Has any single AI neo lab that had a monster seed round due to founder pedigree actually worked out?
English
33
10
220
43K
Ajit Tripathi ری ٹویٹ کیا
emily yu
emily yu@emily_yu·
robotics data for physical AI is front and center this year. from gtc's heavy focus on data infrastructure to human data ecosystems like egoverse, the field is waking up to the bottleneck of scaling robotics data. and there's real divergence in how people think about quantity, quality, modality, and diversity. @aurorafeng_01, @deeptt and i have been looking at the data infra landscape and put together a market map + open data vendor/collector list 👇
emily yu tweet media
English
33
47
330
61.2K
Ajit Tripathi ری ٹویٹ کیا
Wildminder
Wildminder@wildmindai·
NVIDIA says: no more "brute force every pixel" of video understanding. AutoGaze- identifies and removes redundant video patches before they enter a Vision Transformer. Now we can processes 4K long-video in real-time. Works with SigLIP2 and NVILA. autogaze.github.io
English
75
165
2.4K
295.4K
Ajit Tripathi ری ٹویٹ کیا
Andrej Karpathy
Andrej Karpathy@karpathy·
One common issue with personalization in all LLMs is how distracting memory seems to be for the models. A single question from 2 months ago about some topic can keep coming up as some kind of a deep interest of mine with undue mentions in perpetuity. Some kind of trying too hard.
English
1.8K
1.1K
21.2K
2.7M
Ajit Tripathi ری ٹویٹ کیا
Robert Scoble
Robert Scoble@Scobleizer·
I built the most complete list of robotics people and companies here on X (really anywhere): x.com/i/lists/180578… Am watching closely.
English
4
3
27
3.3K
Ajit Tripathi ری ٹویٹ کیا
Cursor
Cursor@cursor_ai·
We're releasing a technical report describing how Composer 2 was trained.
Cursor tweet media
English
167
488
5.1K
1.2M
Ajit Tripathi ری ٹویٹ کیا
Max Weinbach
Max Weinbach@mweinbach·
GPT 5.4 in codex implemented TurboQuant in MLX in like 25 minutes by giving it model weights and the PDF report Sorta insane this is where we are now
Max Weinbach tweet media
English
64
144
2.2K
201.8K
Mitko Vasilev
Mitko Vasilev@iotcoi·
I just implemented Google’s TurboQuant for vLLM. My USB-charger-sized HP ZGX now fits 4,083,072 KV-cache tokens on GB10. This may be the biggest open inference breakthrough of 2026 so far. Training is the flex. Inference is the forever bill.
Mitko Vasilev tweet media
English
70
235
3K
208.2K
Ajit Tripathi ری ٹویٹ کیا
Google Research
Google Research@GoogleResearch·
Introducing TurboQuant: Our new compression algorithm that reduces LLM key-value cache memory by at least 6x and delivers up to 8x speedup, all with zero accuracy loss, redefining AI efficiency. Read the blog to learn how it achieves these results: goo.gle/4bsq2qI
GIF
English
1K
5.8K
39K
19.2M
Ajit Tripathi ری ٹویٹ کیا
Ilir Aliu
Ilir Aliu@IlirAliu_·
Robots building robots. RL token is a simple but powerful idea: Fine-tuning robot policies usually takes days. This takes minutes. Instead of retraining the full model, compress its internal state into a small feature vector and train a tiny RL layer on top. • small actor + critic • trains in real time while the robot practices • adapts specific task stages, not the whole model Result: → precise skills learned in ~15 minutes of robot data This is especially useful for high-precision steps like fastening, screwing, or zip ties. In some cases, it even outperforms human teleoperation in consistency. Big picture: We’re moving from training entire models to patching capabilities in real time. Credit: Physical Intelligence More: pi.website/research/rlt —— if it matters in AI or Robotics you'll read it here first: 22astronauts.com
English
6
37
217
12.9K