lambda

Introducing starscout – a tool that lets you search your github stars using natural language.

@lawrencecchen nice

Introducing cmux Vault: cmux now has a right sidebar with a vault pane where you can see every single Codex, Claude Code, OpenCode, and Pi session You can search through all sessions with full-text search, and drag them directly into your workspace Out in v0.64+

@ClaudeDevs hate this coorporate speech so much, even though i was mainly using claude code cli i will be switching to codex because of this.

Starting June 15, paid Claude plans can claim a dedicated monthly credit for programmatic usage. The credit covers usage of: - Claude Agent SDK - claude -p - Claude Code GitHub Actions - Third-party apps built on the Agent SDK

need someone to built this in rust

Someone tell @AnthropicAI to fix this with mythos.

@trentdotsol @Ledger they have even started forcing firmware upgrade to use the app

@Ledger why can't i upgrade fw from cli? i shouldn't need 150MB worth of bloatware electron app just to keep my device up to date. you still purport to be a security company, no?

I built autoresearch-rl and pointed it at GRPO fine-tuning on @basilic_ai A100s. One command. 15 iterations. Zero human intervention. 100% infrastructure success rate. GSM8K pass@1: 26% baseline to 36%. The hard part wasn't the search algorithm. It was the infrastructure.

Used autoresearch to make @grail_ai GRPO trainer 1.8x faster on a single B200. I kept postponing this for weeks since the bottleneck in our decentralized framework was mainly communication. But after our proposed technique, PULSE, made weight sync 100x faster, the training update itself became the bottleneck. Even with a fully async trainer and inference, a slow trainer kills convergence speed. A task that could've eaten days of my time ran in parallel while I worked on other stuff. Unlike original autoresearch, where each experiment is 5 min, our feedback loop is way longer (10-17 min per epoch + 10-60 minutes of installations and code changes), so I did minimal steering when it was heading in bad directions to avoid burning GPU hours. The agent tried so many things that failed. But, eventually found the wins: Liger kernel, sequence packing, token-budget dynamic batching, and native FA4 via AttentionInterface. 27% to 47% MFU. 16.7 min to 9.2 min per epoch. If you wanna dig deeper or contribute: github.com/tplr-ai/grail We're optimizing everything at the scale of global nodes to make decentralized post-training as fast as centralized ones. Stay tuned for some cool models coming out of this effort. Cheers!

im surprised it took so long for someone to built this

@adiiHQ @ethmumbai was a great session, thank you for doing it!

my session on sales at @ethmumbai

@bajpaiharsh244 happy birthday ser!

Harsh Bajpai is now 24 years old! Which is an interesting feeling tbh …..

@megabyte0x github.com/Beingpax/Voice…

@itzlambda Could you share the link please?

does any local alternative for Wispr Flow exists?

TIL: sqlite has `sqliterc` configuration file Add `.head on` and `.mode column` to ~/.sqliterc. Thank me later.

TIL: sqlite has `ATTACH DATABASE` command. So you dont need import/export scripts to move data between two sqlite databases.

"Any application that can be written in Rust, will eventually be written in Rust."

@bhaxvya @gajesh silicon valley

@itzlambda @gajesh Which show is this

they predicted it bring back this damn genius tv show

@nummanali @LLMJunky oh yes please!

@LLMJunky predicted it $100 Pro plan on the way? Codex will be more accessible Combine 5x Max plan with this and you have everything you need to make all your dreams come true

Code is cheap. Show me the intent.

🤯

OMG Opus 4.6 and GPT 5.3 launch on the same day!