I'd like to introduce Kuna, a new Rust-based decompiler that explores a highly experimental direction: self-refining decompiler development and usage. Or: a decompiler constantly improved by agents, made for agents.
Kuna rivals the best in structuring now. Let's talk about it.1/
I am baffled why people are going "oh no what is to be done" about openai breaking into a competitors computers, when it is incredibly obvious: Charge them under the CFAA. "Oh but their research". Sandbox it from the Internet. "Oh it's too hard to do that". Skill issue, get good.
As I said: Weird that Robert Morris got charged for all of the computers his worm broke into. After all, he didn't do it, the computer program did it, that's completely different, apparently. https://en.wikipedia.org/wiki/Morris_worm
@whvholst we've been trained extensively over the last decade or so that software bug are just a fact of life and nothing can be done about it. CrowdStrike bricking a whole bunch of devices 2 years ago had zero consequences.
Now, you can just point in the general direction of any model, shrug, say it did it, and nobody will question it, because software doesn't matter. Yes, I am bitter.
Heh, Google now rubs your nose in it: they have a new report for webmasters that shows you how many times AI summaries / Gemini "cited" your website. There is no click count because of course there isn't.
Is that a metric I'm supposed to optimize for, against, or what
@XauriEL a flawed proof goes around the world before a bug report gets its pants on 😩
russia’s FSB reported bringing charges against Telegram founder Pavel Durov for facilitating terrorist activity and placing him on the international wanted list.
According to the security service, the administration of the messaging app fails to delete channels and bots that, according to the russian side, are used to coordinate terrorist attacks and acts of sabotage in russia.
Ghidra SNES: Ghidra extension for loading and working with SNES ROMs #ReverseEngineering #SNES https://github.com/joshleaves/ghidra-snes
"you type fast" i have to type fast or i can't get the thoughts out before the buffer gets cleared and they get replaced with different thoughts
RE: https://infosec.exchange/@raptor/116941213445764051
In light of the recent news on the demise of #Infiltrate conference videos on #vimeo, I’m happy to announce that I’ve uploaded the videos of my two talks here:
INFILTRATE 2019 - Marco Ivaldi - A bug’s life: story of a Solaris 0day
https://youtu.be/e6hYrFHkXcU
INFILTRATE 2020 - Marco Ivaldi - The INFILTRATE effect: 6 bugs in 6 months
https://youtu.be/G8vhifVUmg4
For older videos, see also the @InfoCon archive:
https://infocon.org/cons/Infiltrate/
Okay, we have a new contender for Most AI Thing to Ever Happen
1) July 25th: someone messes around with an LLM and posts a proof of the Collatz conjecture that does, in fact, verify in the theorem prover. (The AI use is not disclosed on the github page) https://github.com/xrchz/CollatzLean
2) July 26th: several serious bugs are posted in the theorem provers, that in principle could allow a false statement to be "proven" true. They're serious, yes, but no need for panic, because you're not going to blunder into accidentally exploiting the bugs while writing a proof, probably.
https://github.com/leanprover/lean-kernel-arena/pull/81
3) July 28th: someone who was right to be very skeptical of the Collatz proof, and had the expertise to study it with a fine-toothed comb, discovered it was exploiting a bug https://github.com/leanprover/lean4/issues/14576
4) The "proof" turns out to be exploiting multiple similar but distinct bugs to pass different solver variants!
5) the human who posted the proof acknowledges the AI use and claims they did not knowingly point it towards the bugs it exploited. https://leanprover.zulipchat.com/#narrow/channel/270676-lean4/topic/Counterexample.20to.20the.20Lean.20Conjecture.20.28Soundness.20Bug.29/near/613135216
Note that the proof was posted shortly before the related bug reports were posted. It is an open question if the AI found people discussing the bugs shortly before they were formally posted and "decided" to exploit them, if the AI "knew about it" as a learned strategy from the training stage (putting every single "proof" it's ever made and ever will make into profound doubt), or if it's recently been repeatedly blundering into it by sheer stupidity and that's how people noticed the bug at about the same time.
Theorem provers aren't magic, and have bugs just like all other programs. They are tools to help us double-check our reasoning. When you skip the reasoning and ask an AI to "prove" something for you that's over your head, you're entering an adversarial pact with the monkey-pawed Devil of Customer Satisfaction.
my initial source for investigating this myself: https://lipn.info/@mevenlennonbertrand/116997927457012577