frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Tell HN: ChatGPT and Claude web frontends can run bash commands remotely

1•mabynogy•1h ago
The result for the prompt "run ls on your own server" on ChatGPT.

  total 34K
  drwxr-xr-x  2 root root  160 May  9 17:12 .
  drwxr-xr-x  2 root root  160 May  9 17:12 ..
  -rwxr-xr-x  1 root root    0 May  9 17:12 .dockerenv
  lrwxrwxrwx  1 root root    7 Feb 24  2025 bin -> usr/bin
  drwxr-xr-x  2 root root 4.0K Dec 31  2024 boot
  drwxr-xr-x  4 root root  320 May  9 17:12 dev
  drwxr-xr-x  2 root root   60 May  9 17:12 etc
  drwxr-xr-x  2 root root   60 May  9 17:12 home
  lrwxrwxrwx  1 root root    7 Feb 24  2025 lib -> usr/lib
  lrwxrwxrwx  1 root root    9 Feb 24  2025 lib64 -> usr/lib64
  drwxr-xr-x  2 root root 4.0K Feb 24  2025 media
  drwxr-xr-x  2 root root 4.0K Dec 30 22:28 mnt
  drwxr-xr-x  2 root root 4.0K Jan  8 08:28 openai
  drwxr-xr-x  2 root root 4.0K Jan  8 08:29 opt
  dr-xr-xr-x 20 root root    0 May  9 17:12 proc
  drwx------  2 root root 4.0K Dec 18 04:35 root
  drwxr-xr-x  2 root root  100 May  9 17:12 run
  lrwxrwxrwx  1 root root    8 Feb 24  2025 sbin -> usr/sbin
  drwxr-xr-x  2 root root 4.0K Feb 24  2025 srv
  drwxr-xr-x 12 root root    0 May  9 17:12 sys
  drwxrwxrwt  2 root root  100 May  9 17:12 tmp
  drwxr-xr-x  2 root root 4.0K Feb 24  2025 usr
  drwxr-xr-x  2 root root   80 May  9 17:12 var

Comments

WheelsAtLarge•37m ago
My first thought is that this is a hallucinations. I really doubt anthropic would let something like this go by. Even if it's just local to the user.

Turkey unveils new missile, AI touts as able to hit US mainland

https://www.ft.com/content/d2136091-9fd2-4923-b168-50539e5b27ab
1•JumpCrisscross•21s ago•0 comments

Money is renewable, time isn't

https://get168.app/blog/money-is-renewable-time-isnt
1•dudcks97•27s ago•0 comments

Heuristic Learning

https://trinkle23897.github.io/learning-beyond-gradients/
1•xiaoniu•7m ago•0 comments

Meta's Embrace of A.I. Is Making Its Employees Miserable

https://www.nytimes.com/2026/05/08/technology/meta-ai-employees-miserable.html
2•JumpCrisscross•7m ago•0 comments

Ask HN: Is there evidence that LLMs can extrapolate to new ideas?

1•mikpanko•7m ago•0 comments

Strategic advice from LLM's is "trendslop", say researchers

https://hbr.org/2026/03/researchers-asked-llms-for-strategic-advice-they-got-trendslop-in-return
1•marojejian•8m ago•1 comments

Maybe I'm the Alien [video]

https://www.youtube.com/watch?v=y894m7joEBc
1•hank808•9m ago•0 comments

Speeding Up Zsh

https://arkoinad.com/posts/zsh_speedup.html
1•Brajeshwar•10m ago•0 comments

Trying to Become a Better Programmer by Writing a PHP Interpreter in C++

https://github.com/GiuseppePuleri/jimphp/blob/main/README.md
1•pulerigiuseppe•10m ago•1 comments

The UAP report so far

https://marginalrevolution.com/marginalrevolution/2026/05/the-uap-report-so-far.html
2•paulpauper•11m ago•1 comments

Software Ate My Homework

https://www.theatlantic.com/ideas/2026/05/canvas-hack-campus-fragility/687115/
2•paulpauper•12m ago•0 comments

Jon Rubinstein: Former Apple Hardware Chief on Saving Apple, iMac, iPod, iPhone [video]

https://www.youtube.com/watch?v=PvFMT58lgvk
2•tosh•13m ago•0 comments

Loops and Routines Without Claude

https://ingresslabs.net/blog/spec-driven-development-loops
1•akrylov•13m ago•0 comments

Admit It, That Protein Shake Is Basically Soylent

https://www.theatlantic.com/health/2026/05/soylent-protein-shake/687120/
3•paulpauper•13m ago•0 comments

Cloudflare Pages is down in Spain

https://www.cloudflarestatus.com
2•mariopt•14m ago•1 comments

Getting LLMs Drunk to Find Remote Linux Kernel OOB Writes (and More)

https://heyitsas.im/posts/drinking-llms/
1•hnpufflib•16m ago•0 comments

Mass Gen expands virtual care for 30k patients with no primary care clinicians

https://www.bostonglobe.com/2026/05/09/business/mass-general-brigham-primary/
1•bookofjoe•17m ago•1 comments

Long short-term memory (1997) [pdf]

https://www.bioinf.jku.at/publications/older/2604.pdf
1•andsoitis•17m ago•0 comments

Where will UX design be in 5 years? 5 Predictions (2017)

https://designlab.com/blog/where-ux-design-5-years-predictions
1•downbad_•20m ago•1 comments

I learned something about GPUs today

https://foon.uk/blackshift-sand-bug/
1•birdculture•20m ago•0 comments

Peter Naur – Programming as Theory Building (1985) [pdf]

https://pages.cs.wisc.edu/~remzi/Naur.pdf
1•s4i•21m ago•0 comments

Chicago undergraduate mathematics bibliography (1999)

https://www.ocf.berkeley.edu/~abhishek/chicmath.htm
1•Tomte•23m ago•0 comments

Firefox vs. Chromium security in 2022, how true is it still today?

https://madaidans-insecurities.github.io/firefox-chromium.html
1•wyoh•24m ago•1 comments

GM to pay $12.75M to settle California driver privacy probe

https://www.reuters.com/sustainability/boards-policy-regulation/gm-pay-1275-million-settle-califo...
2•1vuio0pswjnm7•24m ago•0 comments

Show HN: Free C course called "C Systems Lab"

https://csystemslab.com/
2•vhcosta•27m ago•0 comments

Impact of Trump aid cuts: 9 million deaths by 2030

https://www.nytimes.com/2026/05/09/opinion/foreign-aid-cuts.html
7•marojejian•29m ago•1 comments

GM to Pay $13M to Settle Claims It Sold Californians' Driving Data

https://www.law.com/therecorder/2026/05/08/gm-to-pay-13m-to-settle-claims-it-sold-californians-dr...
2•1vuio0pswjnm7•29m ago•0 comments

Show HN: How-to-Train-Your-GPT

https://github.com/raiyanyahya/how-to-train-your-gpt
1•RaiyanYahya•30m ago•0 comments

sick of dealing with unwanted line breaks when you copy text from claude code?

https://github.com/aaronw122/termcopy/
1•aaronw122•30m ago•0 comments

I Hired Too Fast. I Had to Let Everyone Go. Here's What I Learned

https://comuniq.xyz/post?t=1057
2•01-_-•32m ago•0 comments