aaron

Crypto Investor, World Best Algo Trader (Nasdaq, Oil, Gold), SWE

[aaron @pilcloud]

나는 내가 살아온 삶에 만족하는가? - No
나는 내 삶에서 진정으로 나였던 적이 있는가? - No
나는 나 자신에 대해 좋게 느끼고 있었는가? - No
나는 내 삶의 소중한 사람들에게 내 진심을 제대로 표현했는가? - No
내가 죽는다면, 내 영혼이 이 세상에 온 목적을 이루고 죽는 것인가? - No

[밍맹묭 @nextto_mingx2]

[속보]李대통령 "수도권·지방 균형 찾아야…영남보다 호남에 더 비중"https://t.co/MG6nhPi6z2

[개병규 @kyubalnom]

다시 앉으렴

[Polymarket @Polymarket]

JUST IN: Anthropic co-founder Jack Clark reportedly warned new recruits to “get hobbies that aren’t computers,” saying the company is building a “superhuman coder with nation-state hacking capabilities.”

[정민철 @right_mind_2]

이재명의 부정선거라는 광기, 정확하게 팩트체크 해보자고

#이재명 #윤석열 #선거 #투표 #선관위

[널유 @NwolYu_A]

민주주의 위해서 나왔다면서
부르는 노래가 "주님만이 왕이십니다" 이면
대체 어떻게 해석해야함..?

[Alvaro Cintas @dr_cintas]

Google's new algorithm just shrunk 31GB of memory down to 4GB 🤯

TurboVec is a new open-source tool that stores the data your AI app searches through, using 16x less memory.

It runs on Google's TurboQuant, which skips the slow setup step every other tool needs.

→ Faster search than the popular alternative (FAISS)
→ Works on both Mac and standard servers
→ Narrow results to exactly what you want
→ Plugs straight into LangChain and LlamaIndex

Your data never leaves your machine. Runs fully offline, works with Python out of the box.

100% Open Source.

[aaron @pilcloud]

Telegram으로 너무 많은 데이터가 오가는데, hermes를 자체 IM내에서 클플 터널로 통신하려는 게 나으려나.

[aaron @pilcloud]

아니, Hermes 왜 PR을 따로 파는 거야??

https://t.co/QlaN0JbFLe

[β-randonsideshore @pUsAnCetralDong]

스물넷 대학생이 보는 20대 정치성향 1/2

[Kazuki Fujii @kazukifujii]

テックブログ公開 Day5です

FlashAttentionや昨今のHardware Awareな高速化手法を理解したり、提案したりする上で必須となるCUDA Programmingに関して、基礎から解説していくブログシリーズの第一弾です。3万字超えのブログですが、かなり分かりやすく書いていますのでぜひご覧ください。

CUDA Programming Guide Part 1｜Kazuki Fujii https://t.co/5GjsuGIAVS

[Avid @Av1dlive]

https://t.co/iQhxxaTUrb

[Google Gemma @googlegemma]

Meet Gemma 4 12B!

A unified, encoder-free multimodal model designed to bring high-performance intelligence directly to your laptop, and released under an Apache 2.0 license.

Bridging the gap between edge efficiency and advanced reasoning. Here is what’s new with Gemma 4 12B: 👇

[PostHog @posthog]

https://t.co/zqNJaJFxZD

[Logan Thorneloe @loganthorneloe]

Read this to get started learning ML infra.

This is an excellent high-level overview of important considerations in ML training from CMU. It touches on:

- hardware
- memory
- the ML experimentation process

https://t.co/RTWm0Ecni1

[Dan Kornas @DanKornas]

Building coding agents is mostly harness work. This repo shows the pieces.

Dive into Claude Code is a source-level architectural analysis of Claude Code for builders designing AI agent systems.

It helps you move beyond “just call the model” thinking by mapping the runtime around the model: permissions, context management, tool routing, recovery logic, extensibility, and session state.

Key features:

• Source-level breakdown – analyzes Claude Code v2.1.88 across ~1,900 TypeScript files and ~512K lines
• Agent loop map – traces context assembly, model calls, tool dispatch, permission gates, execution, and stop conditions
• Safety + permissions model – documents 7 permission modes, deny-first rule evaluation, and 7 safety layers
• Context management details – explains the CLAUDE.md hierarchy, file-based memory, and 5-stage compaction
• Builder decision guide – turns the analysis into choices for reasoning placement, safety, context, extensibility, subagents, and persistence

It’s open-source (CC BY-NC-SA 4.0 license).

Link in the reply 👇

[aaron @pilcloud]

MSTR-STRC가 넥스트 LUNA-UST일까? https://t.co/2gGZuNn5PJ

[aaron @pilcloud]

재직 중인 모 회사가 국외 뭐 에너지 기업과 엠오유를 맺나 보네. 정부 관계자가 참석해서 축사 예정. 영어 공부 좀 해놔야되겠구만.

근데 전 회사가 더 잘 나가. 엔트로픽 네떡에 합류. 해외 투자가 코 앞이네. 부럽다😭

[aaron @pilcloud]

수천이 날라가네. 이걸 이렇게 담담히 받아들여도 될려나🤧😭

[전목사 @playground777]

야 이거 코인 그냥 개좃되는데.......?
살아날 수가 있냐...?