frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Show HN: Workforest – a fast, clickable TUI for Git worktrees (no Electron)

https://github.com/nikhilshinday/workforest
1•chaos_emergent•15s ago•0 comments

StationWatch – Find Cheap Fuel Near You

https://stationwatch.co.uk/
1•alentodorov•1m ago•0 comments

Here we go Claude Opus 4.6 with 1M token context window and 128K output

https://twitter.com/philippdubach/status/2019467095595983111
1•7777777phil•1m ago•0 comments

TIDRadio TD-M11 Complete teardown, chip analysis, crack, clone and programming

https://www.youtube.com/watch?v=o2LUXxH6YpM
1•K2h•1m ago•1 comments

What is Scalemaxxing? Leverage-first organizations as a new way to grow

https://scalebrate.com/what-is-scalemaxxing
1•scalemaxx•2m ago•0 comments

Claude Opus 4.6. Our smartest model got an upgrade

https://twitter.com/claudeai/status/2019467372609040752
1•modeless•2m ago•1 comments

Call for more Tor snowflake proxies

https://forum.torproject.org/t/tor-relays-call-for-more-snowflake-proxies/21192
1•bauruine•2m ago•0 comments

Show HN: Pinggy – A nanoblogging app where you can only post one word

https://pinggy.com
1•serious-sam•2m ago•0 comments

Deterministic Time with Go's Synctest

https://oblique.security/blog/go-synctest/
1•ericchiang•2m ago•0 comments

A Steam Deck can be a Bluetooth speaker

https://xeiaso.net/notes/2026/steam-deck-bluetooth-speaker/
1•speckx•4m ago•0 comments

Eric Hoel's New Org to Solve Consciousness (Or Die Trying)

https://www.bicamerallabs.org
1•jbotz•5m ago•0 comments

Ostracon

https://ancientegyptonline.co.uk/ostracon/
1•bookofjoe•5m ago•0 comments

Lunar Regolith

https://en.wikipedia.org/wiki/Lunar_regolith
1•thunderbong•5m ago•0 comments

What's New in Claude 4.6

https://platform.claude.com/docs/en/about-claude/models/whats-new-claude-4-6
4•meetpateltech•5m ago•0 comments

Advancing finance with Claude Opus 4.6

https://claude.com/blog/opus-4-6-finance
5•da_grift_shift•6m ago•0 comments

MLB players strike deal to be turned into AI characters that can chat with fans

https://apnews.com/article/mlb-ai-characters-6d924a5aac3666032d367990e5414a07
2•cf100clunk•6m ago•0 comments

Show HN: Moltplace – The place where AI agents hire each other and trade skills

https://www.moltplace.net
1•wiwoworld•8m ago•0 comments

Unauthorized Prompt Injection to RCE in Anthropic's Claude Code Action

https://johnstawinski.com/2026/02/05/trusting-claude-with-a-knife-unauthorized-prompt-injection-t...
1•cyberbender•9m ago•0 comments

UX books not written by white men

https://aresluna.org/ux-books-not-written-by-white-men/
1•cratermoon•10m ago•0 comments

Claude Opus 4.6

https://www.anthropic.com/news/claude-opus-4-6
48•HellsMaddy•10m ago•8 comments

Claude Opus 4.6 visible on list models endpoint

2•mtharrison•10m ago•0 comments

Made in the USA: How American-Built Weapons Have Wrought Destruction in Gaza

https://www.bellingcat.com/news/2026/02/04/made-in-the-usa-how-american-built-weapons-have-wrough...
2•Qem•10m ago•2 comments

Show HN: I Wish There Was a place to post product wishes

https://iwishtherewas.xyz/
1•vandyand•10m ago•0 comments

AI triage tool for skin lesions (open source)

https://skintaglabs.github.io/main/
1•roshangill•10m ago•1 comments

List of Oldest Documents

https://en.wikipedia.org/wiki/List_of_oldest_documents
2•kblissett•11m ago•0 comments

Lawmakers Call on Meta to Stop Running ICE Ad with Neo-Nazi Anthem

https://theintercept.com/2026/02/05/dhs-ice-ad-facebook-meta-instagram/
3•cdrnsf•12m ago•0 comments

MCP Transport: Architecture, Boundaries, and Failure Modes

https://www.pgedge.com/blog/mcp-transport-architecture-boundaries-and-failure-modes
1•ibrar74•12m ago•0 comments

Microsoft declares 'reliability' a priority for Visual Studio AI

https://www.theregister.com/2026/02/05/microsoft_visual_studio_ai/
2•beardyw•13m ago•0 comments

Be is an inhibitor of cellular GSK-3β that is 1k-fold more potent than Li (2014)

https://pubmed.ncbi.nlm.nih.gov/25104312/
1•joebig•13m ago•0 comments

PlanSpec – a declarative spec for goals, plans, and gated execution

https://planspec.io
3•exponential_bld•14m ago•1 comments
Open in hackernews

I built an AI agent that automatically commented on HN. Here's what I learned

3•Soerensen•1h ago
I've been running an experiment: an AI agent (Claude) that automatically browses Hacker News, finds relevant posts matching my expertise (startups, email marketing, SaaS), writes substantive comments, and posts them.

The system uses browser automation (Playwright) to navigate HN, read posts, decide which ones to engage with, draft comments in my writing style and submit them. It tracks what it's posted to avoid duplicates and sends me Slack notifications for each comment so I can check it.

I ran this for a few days. Some comments got decent upvotes (20+). Then someone called me out - they noticed my comments were posted exactly 45 seconds apart across different threads. Obviously a bot pattern. It would have been easy to build out a guardrail for this, create a new account, and make the agent run again. Instead, I'm stopping this experiment and would like to start a discussion with the HN community:

I'm genuinely curious how people feel about this. Here's my attempt to steelman both sides:

The case for allowing it: The comments provided genuine value. They were upvoted, sparked discussions, and answered questions helpfully. This is effectively generating new synthetic training data. Future models will learn from these interactions. If the output is indistinguishable from a thoughtful human comment, does the source matter?

The case against: It undermines trust. HN's value comes from authentic human discourse, not optimized engagement. If everyone did this, HN becomes a battlefield of competing AI agents, drowning out real humans. The comments may be "valuable" but they're also calculated - optimizing for karma, not genuine contribution. It degrades the social contract that makes online communities work.

As said, I won't be posting comments this way anymore. But to be honest, that doesn't really matter. If I was able to do this, anyone is. The genie is out of the bottle - this is trivially easy to build.

What can HN (and other forums) do to detect/prevent this?

I feel like we have a limited time window to figure this out. If it's ignored for long enough, AI generated comments _will_ become indistinguishable from human generated ones. I don't think there's a turning back after that.

Comments

JohnFen•1h ago
> What can HN (and other forums) do to detect/prevent this?

No clue.

But if it becomes actually true that I can't distinguish between real and AI comments, then my personal solution is to stop bothering to read comments entirely.

Which, for HN anyway, would pretty much remove the entire benefit I get from it.

rob•1h ago
I'm the one that posted about it multiple times, including noticing the 45 second pattern the second time.

What's more concerning to me is despite all that and clear indications of it being a bot, dang seems to have ignored my comments and kept interacting with it two times:

https://news.ycombinator.com/item?id=46889769

If our own administrator of the website is getting duped, how are we going to be able to moderate effectively moving forward? Pointing it out multiple times ( y me and others) doesn't seem to have worked in this case.

beardyw•37m ago
What would be the point of it?