Hacker News: alex000kim

New comment by alex000kim in "Research-Driven Agents: What Happens When Your Agent Reads Before It Codes"

alex000kim — Thu, 09 Apr 2026 19:46:02 +0000

That directory is huge already! I guess the index.md helps the agent find what it needs, but even the markdown file is very long - this would consume a ton of tokens.

Also I wonder who/what decides what papers go in there.

In the blog post, the agent is allowed to do its own search.

New comment by alex000kim in "Research-Driven Agents: What Happens When Your Agent Reads Before It Codes"

alex000kim — Thu, 09 Apr 2026 19:32:08 +0000

yup, as the blog says

> The full setup works with any project that has a benchmark and test suite.

so having a clear and measurable verification step is key. Meaning you can't simply give an AI agent a vague goal e.g. "improve the quality of the codebase" because it's too general.

New comment by alex000kim in "Research-Driven Agents: What Happens When Your Agent Reads Before It Codes"

alex000kim — Thu, 09 Apr 2026 19:28:53 +0000

I am sure this would works well in general. There is a challenge wrt to how to make them communicate effectively to e.g. 1) avoid duplicative work and 2) allow them to combine/overlay each others' findings to yield even better results

New comment by alex000kim in "Research-Driven Agents: When an agent reads before it codes"

alex000kim — Thu, 09 Apr 2026 19:26:18 +0000

sounds similar to "LLM Knowledge Bases" https://xcancel.com/karpathy/status/2039805659525644595

New comment by alex000kim in "The Claude Code Source Leak: fake tools, frustration regexes, undercover mode"

alex000kim — Tue, 31 Mar 2026 18:39:51 +0000

technically you're correct, but look at the prompt https://github.com/alex000kim/claude-code/blob/main/src/util...

it's written to _actively_ avoid any signs of AI generated code when "in a PUBLIC/OPEN-SOURCE repository".

Also, it's not about you. Undercover mode only activates for Anthropic employees (it's gated on USER_TYPE === 'ant', which is a build-time flag baked into internal builds).

New comment by alex000kim in "The Claude Code Source Leak: fake tools, frustration regexes, undercover mode"

alex000kim — Tue, 31 Mar 2026 18:11:12 +0000

Oh right, I just saw https://news.ycombinator.com/item?id=47582220 will update the post with this link

The Claude Code Source Leak: fake tools, frustration regexes, undercover mode

alex000kim — Tue, 31 Mar 2026 13:04:30 +0000

Related ongoing thread: Claude Code's source code has been leaked via a map file in their NPM registry - https://news.ycombinator.com/item?id=47584540

Also related: https://www.ccleaks.com

Comments URL: https://news.ycombinator.com/item?id=47586778

Points: 1375

# Comments: 576

New comment by alex000kim in "Migrating from Slurm to Kubernetes"

alex000kim — Wed, 11 Feb 2026 18:51:27 +0000

Author here. I've seen the docs you linked to: Slurm uses "gang scheduling" to mean something specific (timesliced oversubscription where jobs alternate on shared resources).

I'm using the term in its broader CS sense: all-or-nothing co-scheduling of related processes across multiple processors [1]. This is the definition used across the K8s ecosystem e.g. Volcano [2], Kueue [3], and its Coscheduling plugin all define gang scheduling as "all or nothing" allocation.

I still stand by the origianl claim:

Slurm allocates multi-node jobs atomically, while vanilla K8s doesn't. its default scheduler places pods as resources become available, leading to partial allocations and deadlocks for distributed training. It's just a terminology clash. Thanks for the comment anyway.

[1] https://en.wikipedia.org/wiki/Gang_scheduling [2] https://volcano.sh/en/docs/plugins/ [3] https://www.coreweave.com/blog/kueue-a-kubernetes-native-sys...

New comment by alex000kim in "Why Ontario Digital Service couldn't procure '98% safe' LLMs (15M Canadians)"

alex000kim — Mon, 12 Jan 2026 15:43:55 +0000

This was so clearly LLM-generated that I couldn't get through the whole thing.

New comment by alex000kim in "NanoChat – The best ChatGPT that $100 can buy"

alex000kim — Tue, 14 Oct 2025 12:32:28 +0000

I created this PR to make it easier for folks to train and serve it on any cloud (or their own K8s): https://github.com/karpathy/nanochat/pull/18

The Evolution of AI Job Orchestration. Part 1: Running AI Jobs on GPU Neoclouds

alex000kim — Tue, 08 Jul 2025 20:19:42 +0000

Article URL: https://blog.skypilot.co/ai-job-orchestration-pt1-gpu-neoclouds/

Comments URL: https://news.ycombinator.com/item?id=44503694

Points: 2

# Comments: 0

Bulk Object Storage data migration with SkyPilot

alex000kim — Thu, 24 Apr 2025 17:57:13 +0000

Article URL: https://nebius.com/blog/posts/bulk-object-storage-s3-data-migration-with-skypilot

Comments URL: https://news.ycombinator.com/item?id=43785557

Points: 5

# Comments: 1

Orchestrating LLM Fine-Tuning on Kubernetes with SkyPilot and MLflow

alex000kim — Sat, 11 Jan 2025 16:51:21 +0000

Article URL: https://alex000kim.com/tech/2025-01-11-llm-fine-tune-skypilot-mlflow/

Comments URL: https://news.ycombinator.com/item?id=42667086

Points: 2

# Comments: 0

ML experiments in the cloud with Skypilot and DVC

alex000kim — Fri, 11 Aug 2023 01:41:31 +0000

Article URL: https://alex000kim.com/posts/2023-08-10-ml-experiments-in-cloud-skypilot-dvc/

Comments URL: https://news.ycombinator.com/item?id=37084093

Points: 3

# Comments: 0

Why Sales Engineers Exist

alex000kim — Thu, 03 Aug 2023 01:48:15 +0000

Article URL: https://alex000kim.com/posts/2023-08-02-why-sales-engineers-exist/

Comments URL: https://news.ycombinator.com/item?id=36980363

Points: 2

# Comments: 0

Don’t know what to do next? Teach

alex000kim — Tue, 01 Aug 2023 12:58:05 +0000

Article URL: https://alex000kim.com/posts/2023-07-31-teach/

Comments URL: https://news.ycombinator.com/item?id=36955373

Points: 2

# Comments: 0

Show HN: GPT4-powered Slack bot that can scrape URL contents

alex000kim — Mon, 20 Mar 2023 21:10:48 +0000

This repository contains a Python-based Slack GPT Bot that uses OpenAI's GPT model to answer users' questions. Additionally, the bot can extract content from URLs provided in the user's message and take into account their content in its response.

Features - Extract URLs from user messages - Scrape webpage content from URLs - Integrate with OpenAI's GPT-4 to answer questions - Maintain conversation context in a threaded format - Socket mode integration with Slack

Comments URL: https://news.ycombinator.com/item?id=35238787

Points: 8

# Comments: 3

How to Create an Outstanding Data Science Portfolio

alex000kim — Tue, 01 Jun 2021 13:44:19 +0000

Article URL: https://medium.com/practicum-by-yandex/how-to-create-an-outstanding-data-science-portfolio-34f96a66cf4c

Hacker News: alex000kim

New comment by alex000kim in "Research-Driven Agents: What Happens When Your Agent Reads Before It Codes"

New comment by alex000kim in "Research-Driven Agents: What Happens When Your Agent Reads Before It Codes"

New comment by alex000kim in "Research-Driven Agents: What Happens When Your Agent Reads Before It Codes"

New comment by alex000kim in "Research-Driven Agents: When an agent reads before it codes"

New comment by alex000kim in "The Claude Code Source Leak: fake tools, frustration regexes, undercover mode"

New comment by alex000kim in "The Claude Code Source Leak: fake tools, frustration regexes, undercover mode"

The Claude Code Source Leak: fake tools, frustration regexes, undercover mode

New comment by alex000kim in "Migrating from Slurm to Kubernetes"

New comment by alex000kim in "Why Ontario Digital Service couldn't procure '98% safe' LLMs (15M Canadians)"

New comment by alex000kim in "NanoChat – The best ChatGPT that $100 can buy"

The Evolution of AI Job Orchestration. Part 1: Running AI Jobs on GPU Neoclouds

Bulk Object Storage data migration with SkyPilot

Orchestrating LLM Fine-Tuning on Kubernetes with SkyPilot and MLflow

ML experiments in the cloud with Skypilot and DVC

Why Sales Engineers Exist

Don’t know what to do next? Teach

Show HN: GPT4-powered Slack bot that can scrape URL contents

How to Create an Outstanding Data Science Portfolio

Building your Data Science career

New comment by alex000kim in "Fastai: A Layered API for Deep Learning"