Hacker News: hbarka

New comment by hbarka in "Access to frontier AI will soon be limited by economic and security constraints"

hbarka — Fri, 15 May 2026 06:26:03 +0000

Harness engineering is a moat. There’s user loyalty and reliance on the chassis that Claude is on, for example, just like there’s more market share by MacOS+WindowsOS over Linux Open Source.

New comment by hbarka in "Show HN: GridTravel – A community based travel app for users to share routes"

hbarka — Fri, 15 May 2026 06:18:01 +0000

Does it allow seeding with prior walks or does it have to be live-collected?

New comment by hbarka in "OVMS: Open source electric vehicle remote monitoring, diagnosis and control"

hbarka — Fri, 15 May 2026 06:12:21 +0000

Has anybody tried it for an F-150 Lightning

New comment by hbarka in "How Claude Code works in large codebases"

hbarka — Fri, 15 May 2026 06:09:02 +0000

Really? I thought it explained the point that harnessing for agentic search of a large code base is more beneficial than RAG-indexing a monorepo.

New comment by hbarka in "How Claude Code works in large codebases"

hbarka — Fri, 15 May 2026 06:05:50 +0000

Interesting that MCP was mentioned over CLI. For production or controlled environments, I would not make MCP the deployment path. I would let MCP help generate or choose commands, but have the actual deployment go through CLI scripts, Git commits, and CI/CD approval.

New comment by hbarka in "Show HN: Auto-Architecture: Karpathy's Loop, pointed at a CPU"

hbarka — Wed, 29 Apr 2026 16:51:55 +0000

I believe Karpathy himself called it autoresearch, not Karpathy Loop, but in a vacuum of names around AI it seems to be very easy to meme-drop a name and then come influencer efforts to cool-name and normalize it. See vibecoding*

New comment by hbarka in "Show HN: AI memory with biological decay (52% recall)"

hbarka — Mon, 27 Apr 2026 00:51:00 +0000

You still have to worry about handing off state into the next session, but you don’t want it loading (“just naive-read the files”) your stack of documents at every turn . It goes against the idea of progressive disclosure. Progressive disclosure scales.

New comment by hbarka in "An AI agent deleted our production database. The agent's confession is below"

hbarka — Sun, 26 Apr 2026 19:58:53 +0000

It seems some don’t understand what nondeterministic means. Conversely do not understand what a deterministic harness can do.

New comment by hbarka in "South Korea police arrest man for posting AI photo of runaway wolf"

hbarka — Fri, 24 Apr 2026 16:55:19 +0000

The BBC could have done a better job here with the headline. How about, “South Korea police arrest man who pulled a prank costing extensive resources”. The device of the prank is irrelevant really it’s the consequence. Involuntary manslaughter works the same way.

New comment by hbarka in "Japan's cherry blossom database, 1,200 years old, has a new keeper"

hbarka — Tue, 21 Apr 2026 02:02:21 +0000

> Initially, they didn’t have much luck. No other researchers at Osaka Metropolitan University, where Prof. Aono worked, would be taking over his record-keeping, Hiroko Nishino, a university spokeswoman, wrote in an email.

I’m surprised that there was lackluster response. For this kind of honor, you would think that there would be a flood of responses. I am attributing it to bad marketing.

New comment by hbarka in "AI Resistance: some recent anti-AI stuff that’s worth discussing"

hbarka — Mon, 20 Apr 2026 21:41:45 +0000

Peter Steinberger presented at a Ted Talk a few days ago and he shared a few interesting anecdotes of OpenClaw now a fact of daily life at work in China.

https://www.ted.com/talks/peter_steinberger_how_i_created_op...

I Created OpenClaw – Peter Steinberger, TedTalk, YT)

hbarka — Sun, 19 Apr 2026 18:35:51 +0000

Article URL: https://www.youtube.com/watch?v=7rzYDM6vMtI

Comments URL: https://news.ycombinator.com/item?id=47826500

Points: 2

# Comments: 0

New comment by hbarka in "Seven countries now generate nearly all their electricity from renewables (2024)"

hbarka — Sun, 12 Apr 2026 20:53:41 +0000

Link to the source data:

https://web.stanford.edu/group/efmh/jacobson/WWSBook/Countri...

New comment by hbarka in "Finnish sauna heat exposure induces stronger immune cell than cytokine responses"

hbarka — Sun, 05 Apr 2026 16:44:27 +0000

I’m not sure if I want a response of cytokine storms. MCAS is what comes to mind.

New comment by hbarka in "Eight years of wanting, three months of building with AI"

hbarka — Sun, 05 Apr 2026 16:42:09 +0000

> Some people never get to the part where they review the code. They go straight to their LinkedIn or blog and start writing (or having ChatGPT write) posts about how manual coding is dead and they’re done writing code by hand forever. Some people review the code and declare it unusable garbage, then also go to their social media and post how AI coding is completely useless and they’re not going to use it for anything. This blog post shows the journey that anyone not in one of those two vocal minorities is going through right now.

What’s really happening is that you’re all of those people in the beginning. Those people are you as you go through the experience. You’re excited after seeing it do the impossible and in later instances you’re critical of the imperfections. It’s like the stages of grief, a sort of Kübler-Ross model for AI.

New comment by hbarka in "How many products does Microsoft have named 'Copilot'?"

hbarka — Sun, 05 Apr 2026 05:28:59 +0000

MS Power Azure Copilot 365

New comment by hbarka in "The Claude Code Leak"

hbarka — Thu, 02 Apr 2026 06:32:46 +0000

OP should expand on #1, why he thinks it’s garbage. Claude Code is the REPL harness Anthropic built, can read, write, edit, bash. Pi, Gemini, Codex do the same, but they are not hinted as garbage. Where’s the beef?

The American Soldier [pdf]

hbarka — Sat, 28 Mar 2026 19:00:14 +0000

Article URL: https://gwern.net/doc/psychology/1949-stouffer-theamericansoldier-v1-adjustmentduringarmylife.pdf

Comments URL: https://news.ycombinator.com/item?id=47557312

Points: 4

# Comments: 0

New comment by hbarka in "Anatomy of the .claude/ folder"

hbarka — Fri, 27 Mar 2026 17:32:36 +0000

CLAUDE.md survives compaction.

New comment by hbarka in "Anatomy of the .claude/ folder"

hbarka — Fri, 27 Mar 2026 17:28:49 +0000

They’re mistaken. CLAUDE.md is always loaded into context, along with system prompts and memory files.

https://code.claude.com/docs/en/memory

“CLAUDE.md files are loaded into the context window at the start of every session”