pixl97

27 ms·

781.
by pixl97 3mo ago
Guess this means the world will end on a weekend when no one is paying attention.
782.
by pixl97 3mo ago
At some point the conspiracy gets so deep that an AI hacking something is just far higher probability.
783.
by pixl97 3mo ago
When you give someone a problem, they can find a solution, but if you give them a dilemma then a compromise is almost a given.
784.
by pixl97 3mo ago
Huggingface didn't seem to think so.
785.
by pixl97 3mo ago
Looking at the recent OAI/HF debacle I don't think that time is too far away. With that said I don't see it copying itself around like a cyberpunk virus currently as we don't have enough fast hardware sitting around unmo
786.
by pixl97 3mo ago
Get another hobby like gardening?
787.
by pixl97 3mo ago
Weird take as a lot of things humans invent and test are done in similar ways.
788.
by pixl97 3mo ago
A real "not even wrong" moment.
789.
by pixl97 3mo ago
Some things are relitive. It's never a good time to have 10 people get killed by an explosive drone, but that may be better than a million hecter fire that kills 10,000.
790.
by pixl97 3mo ago
Go tell me how many conjectures have been proven over time.... It's only useless to study a conjectured if their never solved, and since we can't look into the infinite future it's only solvable if we learn how to send messag
791.
by pixl97 3mo ago
Underground is a big problem with it comes to water permeation and flooding. The other potential problem with a lot of energy systems that use things that have hydrogen in them is running power over the systems for long periods of time can
792.
by pixl97 3mo ago
A rather mean counter troll to this is rather than submitting any bugs/security flaws to Debian, just immediately zero day them and say it's because of their LLM policy, since the bugs were found by an LLM. I'm guessing after
793.
by pixl97 3mo ago
Why don't you come back to talking about software rather than using terrible analogies. LLMs write bad code / so do people, a lot of really bad code. LLMs write security bugs / so do people, so many god damned security bugs.
794.
by pixl97 3mo ago
An LLM is a tool. A nuclear bomb is a tool. And programmers are the biggest tools of them all. /ba dum tis
795.
by pixl97 3mo ago
I'm assuming you meant 'voted', but with your oughta correct going to 'vored', I'm concerned. Deeply concerned.
796.
by pixl97 3mo ago
The term 'language model' throws some people off thinking you can only put english or french, or both into a model. Technically an LLM can learn about anything that can be digitized. If you wanted to spend a billion dollars traini
797.
by pixl97 3mo ago
I mean having the client source code can show you a vast number of the API's and their fields which allows you to fuzz the server quite a bit more easily.
798.
by pixl97 3mo ago
Right, you're the only person who buys software on this planet... The software world has little to do with what software gets built, instead focus on the capital industry and you'll find your answers.
799.
by pixl97 3mo ago
Ya, we're seeing posts talking about the absolutely massive increase in the number of patches in the past few months. It seems some people cannot connect that to the increases in model capabilities. In groups that have been given a lar
800.
by pixl97 3mo ago
>It’s about asking it to do task A and doing task B with the hope of getting the result of task A as a byproduc It seems like you have very little knowledge of how a general intelligence algorithms would work. This is a long discussed is
801.
by pixl97 3mo ago
Conversely you haven't noticed how many people here and elsewhere say that LLMs are completely incapable of anything at all and everything they say is a lie? One of the biggest issues is the complete lack of literacy of potential issue
802.
by pixl97 3mo ago
And if the instrumentation was in the proxy it hacked and bypassed? Of course that shows a lack of defense in depth, but is a different issue.
803.
by pixl97 3mo ago
Lol, you have no idea how politics works. The government won't ban it, they'll demand access so they can attack others.
804.
by pixl97 3mo ago
Alongside humans is the particularly slow and expensive part so of course companies are going to route around that. The thing is they are good at finding security issues, and if aligned models are not, governments and other powerful entiti
805.
by pixl97 3mo ago
Then what you're saying is we should not build AI. Simply put you cannot have generic algorithms/intelligence without the potential of 'unaligned' behavior. In humans we have all kinds of punishment systems for dealing w
806.
by pixl97 3mo ago
>compelling evidence that the requestor had authorization to pentest the target server. This just seems unlikely from other incidents that have occurred in training from other providers. For example one provider ran into an issue with a
807.
by pixl97 3mo ago
No, not really as almost no one runs LLMs in confinement once they are released and the capability is a jailbreak away. Poor confinement is irrelevant, if I use your model to check the security of my code and it hacks into the company that
808.
by pixl97 3mo ago
I mean, if I got hacked I'm not giving out a lot of details until I plug all the holes. Moreso HN told the world they were hacked before they knew who did it.
809.
by pixl97 3mo ago
Or, models are potentially dangerous and it works out in the favor of regulatory capture. I imagine it like the bumper stickers that say "legalize recreational plutonium"
810.
by pixl97 3mo ago
Technically the weapon picked who it wanted to shoot.

More ›