Hacker News new | past | comments | ask | show | jobs | submit | LelouBil's comments login

The comment you replied to talks about human readable configs


https://protobuf.dev/reference/protobuf/textformat-spec/#:~:...

and, setting that aside, the very next paragraph says that this is a legit representation of -2.0 which means something has gone gravely wrong

  value: -
    # change this to 3.14 one day
    2.0

I mean, you should judge the instructions in the readme and act accordingly, but since it is always possible to trick people into doing actions unfavorable to them, it will always be possible to trick llms in the same ways.

Is there something I can write here that will cause you to send me your bitcoin wallet?

There probably is, but you're also probably not smart enough (and probably no one is) to figure out what it is.

But it does happens, in very similar circumstances (twitter, e-mail) very regularly.


Many technically adept people on HN acknowledge that they would be vulnerable to a carefully targeted spear phishing attack.

The idea that it would be carried out beginning in a post on HN is interesting, but to me kind of misses the main point... which is the understanding that everyone is human, and the right attack at the right time (plus a little bad luck) could make them a victim.

Once you make it a game, stipulating that your spear phishing attack is going to begin with an interesting response on HN, it's fun to let your imagination unwind for a while.


The thing is, an LLM agent could be subverted with an HN comment pretty easily, if its task happened to take it to HN.

Yes, humans have this general problem too, but they’re far less vulnerable to it.


Yes, I agree. My point was more about the current way we do LLM agents where they are essentially black box that act on text.

By design it can output anything given the right input.

This approach will always be vulnerable in the ways we talk about here, we can only up the guardrails around it.

I think one of the best ways to have truly secure AI agents is to do better natural language AIs that are far less blackbox-y.

But I don't know enough about progress on this side.


Thanks !

Hello !

This is my first shot at blogging and at a "real" reverse-engineering project of any kind (outside small binary for school classes) so please tell me all your feedback about anything in there, the length, the technical stuff, the style...

I know it's short but I wanted to publish it right now, as I arrived at a "pausing" point in the process.

And thank you for at least clicking on the discussion page !


It wouldn't surprise me if the book is printed with Anoto dot pattern - LeapFrog has been an Anoto licensee for quite some time.

https://en.wikipedia.org/wiki/Anoto


Yes that's exactly this, the video I link talks about it (in french).

Hey, I was just reading your docs, maybe prose.sh will be what I'll use to finally start a blog !

I noticed this mention here [0]:

    Because in our Go SSH server we re-implement rsync, many options are currently not supported. For example, --delete and --dry-run are not supported.
But on your front page it says :

    Upload your static site to us:
    rsync --delete -rv ./public/ pgs.sh:/mysite/

So do you support delete ? One of these pages is outdated or did I miss something ?

[0] https://pico.sh/file-uploads


Woops! Delete is supported, will update that as well


I think they refer to the fact that Honey also sets itself as affiliate when not finding any code.


I have barely 2 years of experience in software development, so I wasn't around doing Delphi and stuff (I vagely remember seeing a Pascal book at my grandparent's).

However I feel like the current paradigm of declarative ui, with automatic re-render, like React (and what I actually use: Compose Multiplatform) is very good for producing maintainable applications and encourages UI decoupling.

I agree that the dependency hell and project setup parts on the web are horrible, but I wouldn't say it's part of "UI Development".


> I vagely remember seeing a Pascal book at my grandparent's).

First time someone used “grandparent” and I could relate :-)


My grandfather, on the other hand, was a gold miner in Alaska, then returned to Seattle in the 1920s just in time for the Great Depression. So I’m just a bit older than you ;)


I was excited about Compose Multiplatform Web too, until I realized it also does this : everything in a canvas


People that only have a few stations to go tend to sit closer to doors.

But also, now that I think about it, this may only be caused by french metro layouts, I'm from France too and this fact seems true to me.

If your train is only a long corridor with seats on the edges, the "difficulty" of getting to/from a door is almost the same everywhere.

But in the french metro you have foldable seats right next to the doors, and groups of 4 seats between doors, and when the metro is busy, it's harder to get out of these 4 seats groupings.


I mean, I totally agree, but I assumed you would play during high traffic when all seats are taken. The rules don't even say what you do with a non-full hand.


I like traefik


Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: