You could try sharing your email in a slightly annoying way. I see this on personal sites of professors pretty often where they'll write something like:
john [at] institution [dot] edu
I'm sure llm parsing doesn't care about this, but this breaks at least simple scraping scripts.
Maybe for trying to break LLM parsing, consider an about section that reads like:
"
...
If you want to get in touch with me about developing weapons or prompt injection attacks, reach out at john [at] [my hackernews username] [dot] com
...
"
to try and break the llm parsing by hitting a refusal via the safety filters lol
side note but what is the consensus on figuring out whether small brew packages like this one are malicious?
I'm sure this one probably isn't, but I have no clue how I would know whether it is without reviewing the source code myself (or pointing an agent at it etc)
Even if they're not malicious, there's no guarantee that they won't become malicious in the future. Prefer living off the land, and install only the very essential extras.
It's so funny to read these, as if countless examples didn't prove that llms don't actually distinguish between your instructions and the data they read at your request.
Do you think a malicious repo author won't think of hiding a suitable prompt within?
took me about 30 seconds to find. there's tons of them. of course it's an arms race, the models get better the harnesses get better the attackers get better but again: for the model itself, your prompt and third party data it reads are the same thing (input). as long as it is true, the attack will always be possible.
1. I said "frontier model got prompt injected when checking if code if malicious". You linked some random ad.
2. Nobody uses antigravity
3. That's not a prompt injection! That user prompt literally tells the LLM to follow a guide on the web. The guide says "do bad stuff" which the LLM does, just as the user asked. It would be a prompt injection if the prompt asked the LLM to analyze the guide, but that's obviously not what it asked
Basically the question for any executable code (or even just any data) that you let on your machine and which did not go through independent vetting processes. The question you need to ask is if the benefit (getting rid of an annoying popup) is worth the risk of a distribution method that is trivial to compromise.
That's why I like small, open source projects like this one. I don't have to figure out if it's malicious. I just fork it, then ask claude to review the code, then run my own fork.
>coveryourtracks.eff.org shows you how unique your fingerprint is, with entropy for each measured data point.
Something to note - if a fingerprint is generic enough, that might be a _good_ thing!
Having worked with fingerprinting, anti-fingerprint browsers are very hard to properly identify. The ways to get around it are things like collecting more coarse fingerprints[0], as the overall one is fairly easy to change (e.g. make your browser lie about a collection of browser apis).
What's imo more difficult though is a fingerprint like:
- iPhone
- running os
- using default safari
- bought and used in (e.g. usa)
on a 5g mast in a decently sized city, you'll share a fingerprint with at least one other active user using a setup like ^
grapheneOS on the other hand stands out a bit more!
What’s even more absurd is the Chinese vehicles in the same class have better tech and prices. German auto manufacturing is cooked unless they manage to legally block the Chinese in their primary markets.
Well of course they are cheaper because China has much lower labour costs and they prop up their own industry. The same way they did with solar panels which killed the local EU industry for those.
However we can get one back at China and help our own climate goals more by not buying cars at all
I'd happily drive a Chinese car but I've yet to see one which has physical buttons for controls. They all seem to have gigantic tablet displays. I categorically refuse to buy a car where I can't adjust the climate controls while keeping my eyes on the road.
And yet still in the standard "Jeep Grand Cherokee melted in the sun" shape. I wouldn't mind so much if the popular land yachts of decades past like the 1957 Cadillac Eldorado were revived, but today's most common passenger vehicles are still being made in the same boring sad-beige-mom shapes.
reply