Mahesh Sathiamoorthy

4.5K posts

Mahesh Sathiamoorthy banner
Mahesh Sathiamoorthy

Mahesh Sathiamoorthy

@madiator

RL Environment Curation. Data Curation (OpenThoughts). Post-training. CEO @bespokelabsai. Ex-GoogleDeepMind.

Inside a RL Environment Katılım Şubat 2008
1.4K Takip Edilen14.7K Takipçiler
Sabitlenmiş Tweet
Mahesh Sathiamoorthy
Mahesh Sathiamoorthy@madiator·
We are announcing Open Thoughts, our large-scale open-source effort to curate the best open reasoning datasets! DeepSeek-R1 is amazing but we still don't have access to high-quality open reasoning datasets. These datasets are crucial if you want to build your reasoning models! Bespoke Labs released a 17k reasoning dataset last Wednesday, and the reception has been phenomenal (it's trending on HF). So we are joining forces with the Datacomp community to launch Open Thoughts --- an open data, open model, and open code initiative for creating the best open reasoning datasets and the associated models. Along with this, we release OpenThoughts-114k reasoning dataset and the associated OpenThinker-7B model. Links to the code, model, and data are below in 🧵.
Mahesh Sathiamoorthy tweet media
English
46
288
1.8K
229.3K
Horace He
Horace He@cHHillee·
While I'm happy that many folks seemed to enjoy this talk, there are a lot of inaccuracies in this tweet 😆 "Jane Street hired" - I've never worked at Jane Street "This junior" - at this point I'm 5 years out of undergrad, so I think arguably I'm not a junior anymore although perhaps some would disagree :) "uses AI to analyze ... data" - I would not describe my role like this haha Probably also good to mention that it's from the Jane Street Tech Talk series: youtu.be/139UPjoq7Kw?si… and not from this reposter
YouTube video
YouTube
bodila@51bodila

Jane Street hired this junior at $220k-$600k /year because he uses AI to analyse TRILLIONS of data in this 1-hour lecture - he show how to research trillion of data points thanks to his machine Bookmark & watch it, instead of Netflix to learn how to do the same!

English
34
93
2.1K
357.5K
Bespoke Labs
Bespoke Labs@bespokelabsai·
We are excited to welcome Avinash Arjavalingam as a Member of Technical Staff at Bespoke Labs. In his previous role, Avinash was a Software Engineer at LinkedIn working on their Relational Databases team. He also holds a Masters and Bachelors degree in computer science from UC Berkeley, specializing in databases and distributed systems.
Bespoke Labs tweet media
English
2
2
25
3.9K
Mahesh Sathiamoorthy retweetledi
Han Xiao
Han Xiao@hxiao·
bro goes casual mode at ICLR
Han Xiao tweet media
English
11
24
522
42.8K
Yu Bai
Yu Bai@yubai01·
🥔GPT-5.5 is here in Codex and ChatGPT. 🚀 Don’t want to keep saying “step change” with each release, but this time we feel it’s a pretty big one. It may be an inflection point for a lot of things down the road. Please try using this model in Codex for your coding and other professional use cases – Start with the same tasks as before, expect it to do more with less human in the loop, it will make a big difference over 5.4.
OpenAI@OpenAI

Introducing GPT-5.5 A new class of intelligence for real work and powering agents, built to understand complex goals, use tools, check its work, and carry more tasks through to completion. It marks a new way of getting computer work done. Now available in ChatGPT and Codex.

English
2
2
44
2.4K
Mahesh Sathiamoorthy
Mahesh Sathiamoorthy@madiator·
Jensen mentions converting electrons to tokens. But I go one step further to say that humanity's crowning glory is the achievement of converting electrons to intelligence. So much science, art, and perseverance that has gone into doing it. And ultimately this intelligence will change the universe by rearranging the matter and electrons.
Dwarkesh Patel@dwarkesh_sp

The Jensen Huang episode. 0:00:00 – Is Nvidia’s biggest moat its grip on scarce supply chains? 0:16:25 – Will TPUs break Nvidia’s hold on AI compute? 0:41:06 – Why doesn’t Nvidia become a hyperscaler? 0:57:36 – Should we be selling AI chips to China? 1:35:06 – Why doesn’t Nvidia make multiple different chip architectures? Look up Dwarkesh Podcast on YouTube, Apple Podcasts, Spotify, etc. Enjoy!

English
1
0
2
1.6K
Mahesh Sathiamoorthy
Mahesh Sathiamoorthy@madiator·
@deedydas Ending was cut off, so let me finish it: she realizes how hard it is to shoot with the camera and whips out claude code + seedance 2 api and creates this exact video. Videoception.
English
0
0
7
340
Deedy
Deedy@deedydas·
i generated this entire 45s movie clip (audio + video) with claude code + seedance 2 api there's still telltale AI smell, but we should be at full length movies indistinguishable from real ones by the end of the year (veo 5)
English
173
94
1.4K
133.7K
Mahesh Sathiamoorthy
Mahesh Sathiamoorthy@madiator·
Buckle up everyone, your API costs are going up, not down.
Mahesh Sathiamoorthy tweet media
English
0
1
5
430
Mahesh Sathiamoorthy retweetledi
Tanay Padhi
Tanay Padhi@tanaypadhi·
how did Allbirds pivot to AI compute hardware before the shoe company literally called ASICS
English
69
492
7.3K
348K
Ollie Liu
Ollie Liu@olliezliu·
I recently joined @reflection_ai to work on safety and alignment post-training. Open-weight superintelligence presents new constraints and technical challenges in safety, and I'm incredibly grateful to be working with an exceptional team. Separately-just moved to SF. Would love to see friends in the Bay :-)
Ollie Liu tweet media
English
17
3
147
10.2K
Mahesh Sathiamoorthy
Mahesh Sathiamoorthy@madiator·
I want to have a version of claude app running strictly locally with all sorts of my sensitive info stored in the memory. I am filling some forms and need to enter the passport number, expiry date, and what not. It will be nice to have it all in one place in a local model+agent which, at the very least gives me those info when I ask (rather than me having to navigate to various folders, and opening files to glean that info), or it should be able to fill those forms.
English
0
0
4
843
Mahesh Sathiamoorthy retweetledi
Bespoke Labs
Bespoke Labs@bespokelabsai·
Great validation for Bespoke Labs Chief Science Officer, @AlexGDimakis's work. Anthropic's work appears to be based on the Advisor model paper [1, 2] by Alex and co-authors. As a reminder, we at Bespoke are hiring top researchers to work on RL and RL Environments and other cool stuff [3]! Links in the comment!
Claude@claudeai

We're bringing the advisor strategy to the Claude Platform. Pair Opus as an advisor with Sonnet or Haiku as an executor, and get near Opus-level intelligence in your agents at a fraction of the cost.

English
2
3
27
3.7K
Mahesh Sathiamoorthy
Mahesh Sathiamoorthy@madiator·
Oh yeah, I almost forgot that Boris had left to Cursor for a brief month or so last year. Wonder how the world will look like had he stayed there!
Mahesh Sathiamoorthy tweet media
English
1
1
74
10.4K
Mahesh Sathiamoorthy retweetledi
Alexandr Wang
Alexandr Wang@alexandr_wang·
1/ today we're releasing muse spark, the first model from MSL. nine months ago we rebuilt our ai stack from scratch. new infrastructure, new architecture, new data pipelines. muse spark is the result of that work, and now it powers meta ai. 🧵
Alexandr Wang tweet media
English
727
1.2K
10.3K
4.5M
Jiahui Yu
Jiahui Yu@jhyuxm·
Happy to share Muse Spark, a natively multimodal reasoning model w/ tool-use, visual chain of thought, and multi-agent orchestration! It’s been a fulfilling journey not just building the model, but the team and culture behind it. Now live in product. ai.meta.com/blog/introduci…
Jiahui Yu tweet media
English
20
52
447
41.9K
Mahesh Sathiamoorthy
Mahesh Sathiamoorthy@madiator·
Wonder if there was any previous tech cycle that had made people this excited!
English
3
0
12
2.5K