frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Ask HN: If AI tools are so good, why isn't your bug backlog empty?

90•ex-aws-dude•1d ago•108 comments

Ask HN: UK Equivalent for Power Switch Tail/IoT Power Relay?

2•a2fz•5h ago•0 comments

Ask HN: Are We in a Cyberwar?

9•bhartzer•7h ago•17 comments

Ask HN: Anyone else blurting?

7•l33tbro•6h ago•5 comments

Ask HN: What was your weirdest or funniest job rejection?

7•cristaloleg•18h ago•14 comments

Ask HN: Are the more software malfunctions now than before?

10•sirnicolaz•1d ago•6 comments

Ask HN: Are iPhone Duo Customers Happy?

16•rogerdickey•1d ago•16 comments

Ask HN: What are you working on? (November 2026)

305•david927•1w ago•1106 comments

Ask HN: Syncthing

7•bikvfeee•18h ago•5 comments

Ask HN: What do you think is going on with that many outages coming in a row?

9•vld_chk•1d ago•10 comments

AI is shortening the time from PHP flaw disclosure to working exploit, Microsoft

3•codeswithrocky•8h ago•0 comments

Ask HN: Where are you on the spectrum of AI/LLM adoption?

4•joddystreet•21h ago•7 comments

Tell HN: HN is being scraped for emails to scam

122•eks391•2d ago•42 comments

Is the Economy Over-Indexed to Software?

5•nizsle•1d ago•6 comments

Ask HN: How realistic is an LLM-powered worm that spreads from PC to PC?

3•anon115•6h ago•3 comments

ChatGPT Desktop forwards your SSH agent to remote machines

222•DenGorbachev•1w ago•80 comments

Ask HN: What does building hardware with AI look like?

9•mjs06•2d ago•3 comments

Tell HN: Google Search to start locking down search behind a login wall

47•exploraz•4d ago•30 comments

Ask HN: What is the state of self-hosting email in 2026?

48•0o777•4d ago•36 comments

Ask HN: Do you still enjoy coding in C++?

67•hasley•1w ago•92 comments

Ask HN: Most compelling early stage fund pitches?

5•bix6•1d ago•5 comments

Ask HN: What is a better name for Claude Code?

3•nstj•1d ago•3 comments

AI is killing cheap VPS servers/manageable pricing

13•SudoSH•22h ago•14 comments

You've reached the end!

Open in hackernews

Ask HN: How realistic is an LLM-powered worm that spreads from PC to PC?

3•anon115•6h ago

Comments

strangecasts•5h ago
Anthropic had a very hotly debated proof-of-concept last month: https://news.ycombinator.com/item?id=50061603

This proof-of-concept skipped the work of exploitation to start running the code, assuming that there are security holes that make IoT/smart TVs/older devices essentially exploitable "for free", and just focused on the worm/replication [1] part. They were able to have a 27B model run sharded across an impromptu network of set-top boxes, but also showed that this wasn't particularly practical - the resulting network output less than a token a second.

This suggests that you probably won't see "LLM-powered worms" in the sense that runs an LLM on each infected device (which becomes noticeable in terms of power consumption and/or temperature), but rather LLMs as additions to typical client-server botnets, where the models run on more powerful servers and manage infected devices, using them to probe other devices on the network and stage attacks.

[1] I am a bit annoyed that "self-replication" is the term they use, since the proof-of-concept requires the harness to have access to the weights - if you allow that then you can make a <100-parameter "self-replicating" model by outputting the tokens "please run `cp skynet.safetensors skynet2.safetensors`"...

mrlonglong•5h ago
Pfft.

You'd need massive amounts of memory and bandwidth to replicate.

One day maybe, but now? No.

muzani•4m ago
It's an abstract concept until the first person does it. I don't think it's impossible, evolutionary programming would be very interesting here. I'd give it a shot if they didn't arrest people for trying things like this.