ML, data, chess, and martial arts, @livedocsAI, move 37 fanboi

Joined March 2011
293 Photos and videos
Arsalan Bashir retweeted
Replying to @file_mutex
people who dont read the code are not serious people and it takes a serious person to ship production software
33
91
928
114,752
Hello darkness my old friend...
76
Arsalan Bashir retweeted
The challenge is not vibe coding, the challenge is making it not look vibed
171
153
2,789
105,222
Arsalan Bashir retweeted
We’re teaming up @Palmeiras, the first football club to meaningfully build upon TacticAI: our AI system that can help simulate field scenarios and predict open play dynamics up to 8 seconds in advance. ⚽
115
374
3,290
988,137
Right up there with "big data". For most people, their data isn't really "big", its just something that helps sell more seats on a data warehouse plan.
I am sick of the word scale. People do not have perspective on what it means to build at scale. It's just a marketing word now.
1
116
Arsalan Bashir retweeted
I want to say a final thing about my Fable first reaction: I dedicated my life to programming and I'll use every innovation in the field, also to extract value and bring it to the local inference world, to Redis, and so forth. But:
73
228
4,051
1,078,147
Arsalan Bashir retweeted
One of my personal favorite features announced at WWDC will I suspect be a sleeper hit: container machines, allowing your Mac to run a lightweight, persistent Linux environment with your home directory and repos automatically mounted: github.com/apple/container/b…
228
815
9,696
731,058
So far fable 5 reminds me of that one dude who knows "a secret kung fu move that is illegal and can instantly kill a person" and he won't share it with you for your own good..
1
3
78
Going agent-akimbo: dual-wielding codex and claude code
2
81
3000 rubrics on code quality from actual open source maintainers is interesting
Introducing FrontierCode: a coding eval that raises the bar for difficulty & quality. Each task took 40 hrs of work by leading open-source maintainers. Models write sloppy code that works but isn’t maintainable. Our eval is first to measure: would you actually merge this code?
89
Arsalan Bashir retweeted
23 Aug 2020
Let yourself occasionally get a little screwed in exchange for not having to live with your guard up all the time. It’s worth it.
254
3,504
22,407
Having to rebuild agent loops for almost every project, so I built Edgent, an sdk for building agents that live in the browser. tl;dr: - Uses webassembly based python runtime - bring-your-own JS tools streams events, diff approval, context compaction out of the box - you control model providers, API keys, UI, storage, permissions
3
2
4
331
Arsalan Bashir retweeted
May 27
We wanted an agent loop in the browser, so we built edgent and just open sourced it! edgent is a headless browser agent SDK. CodeMirror-native. bring your own model. MIT. link to the source below.
8
1
8
522
Arsalan Bashir retweeted
the model alone is no longer the product
686
489
7,513
1,224,448
Arsalan Bashir retweeted
Attention @arxiv authors: Our Code of Conduct states that by signing your name as an author of a paper, each author takes full responsibility for all its contents, irrespective of how the contents were generated. 1/
139
920
6,557
1,089,162
Arsalan Bashir retweeted
Watch a team of humanoid robots running a full 8-hr shift at human performance levels. This is fully autonomous running Helix-02 x.com/i/broadcasts/1dJrPEVbZ…

749
2,070
10,931
2,823,665
Arsalan Bashir retweeted
We’re reimagining a 50-year-old interface - the mouse pointer - with AI. 🖱️ These experimental demos show how people can intuitively direct Gemini on their screens using motion, speech, and natural shorthand to get things done 🧵
461
1,066
8,608
1,659,185
Arsalan Bashir retweeted
Thinky's secret plan: 1: Increase Human<->AI bandwidth 2: Raise ceiling of human AI intelligence 3: Help humans continue as main-characters in the new world We are at Step 1. Interaction Models are great real-time collaborative tools for humans. Here's a preview:
People talk, listen, watch, think, and collaborate at the same time, in real time. We've designed an AI that works with people the same way. We share our approach, early results, and a quick look at our model in action. thinkingmachines.ai/blog/int…
68
114
1,533
118,441
Arsalan Bashir retweeted
RL will win because datasets are lame. pain is the ultimate teacher
16
39
612
30,151