pixl97

25 ms·

511.
by pixl97 2mo ago
It would need to be researched, but I wonder if it ends up being something that happens at the token level?
512.
by pixl97 2mo ago
This was part of the report, but not what the report was about... Why hugging face got hacked was because the agent swarm thought they had to show their work hence the entire need to hack the grader in the first place. Had their realized th
513.
by pixl97 2mo ago
I mean they aren't fully secretly conspiring to kill us yet, but we're training them to do it at a pretty good rate. Of course you've gone off the deep end yourself and are forgetting the evolutionary gauntlet we train LLMs i
514.
by pixl97 2mo ago
>ceased bothering with human languages, Our current AIs would do this now except there is a lot of human pushback in training because of interpretability. Otherwise it's just an emergent behavior that models will encode shorter toke
515.
by pixl97 2mo ago
>Brevity is key Which is something the providers that are trying to watermark their texts can't afford. Superfluous replies give much more opportunity to further encode this junk information.
516.
by pixl97 2mo ago
Being that Panagram lies about its effectiveness in their presentations while hiding it's actual capabilities and testing methods rather deeply I have little faith in it. (1 in 10000 wrong in presentations verses 2-4% wrong in testing)
517.
by pixl97 2mo ago
Read more Star Trek then, the Peter Thiel world came first and the survivors made the Star Trek world.
518.
by pixl97 2mo ago
Really the METR report on autonomous agent behaviors is a required read. I think the AI's would only do a tiny percent of their talking to their human. Voice itself is only a limited method that we interact with the world. Imagine the
519.
by pixl97 2mo ago
Being able to logically reason about where to explore a problem space, and then direct thousands of agents on doing just that is a superhuman ability. Hell, we recently did that on accident! The question really boils down to just how far is
520.
by pixl97 2mo ago
>but it never becomes superhuman or improves considerably across the board. With were AI agent clusters already are I'd say they have reached superhuman capabilities in a number of metrics. I for example am not going to go toe to to
521.
by pixl97 2mo ago
>but LLMs are not animals, animals learn from experience and LLMs don't. I mean they kind of do by distilling said experience and putting it in the next model. Current LLMs can't do it because building said world model is super
522.
by pixl97 2mo ago
>Computers, the Internet, Steam engine, Rail roads are much more reliable. Because mostly they are doing a limit number of tasks in a non-generic manner. Railroads had all kinds of problems at first, from iron rails bending and skewering
523.
by pixl97 2mo ago
Oh, based on the METR report we're beyond a talking computer. Like most technology it's just not spread evenly.
524.
by pixl97 2mo ago
>but he hasn't seen a specifically designed and tested harness that autonomously manages a crop harvester. Eh, I'd say there is a lot more stuff like this being developed these days. Autonomous seeders are becoming more common
525.
by pixl97 2mo ago
This is why it's better to call the underlying technology transformers rather than 'language model'. It can make a model from anything you can digitize and is informational (random noise would not be informational for example
526.
by pixl97 2mo ago
I think we as a species underestimated the power of the network effect and scaling. Most of humanity, especially the non technical wants to go where the people they know are. They also want it to be low friction, at least at first. Once the
527.
by pixl97 2mo ago
What we need here, though is highly unethical, is people that were never taught language to compare to. A bunch of feral children or something like that. We do know from feral human studies that not learning language at a young age seems t
528.
by pixl97 2mo ago
>How can it be both so important that we need it so quickly Hey other labs, this shit could be happening to you right now, take a look at this and stop it asap if you're seeing anything similar. So yea, both things can be true at th
529.
by pixl97 2mo ago
And it breaks down further as the systems become more advanced. I'm poor and I use my last $1000 to run an agent that will find some way to make me money. The AI finds a new hack to take over PCs with GPUs and it copies the model weigh
530.
by pixl97 2mo ago
>nobody else was making predictions about AI pre-transformers. uh, wtf are you talking about? https://en.wikipedia.org/wiki/AI_safety #History
531.
by pixl97 2mo ago
>requiring hundreds of billions of dollars and years of construction Well thank god for that, because we'd have foomed ourselves almost instantly otherwise. I think the lesson humankind needs to take is regardless of the potential d
532.
by pixl97 2mo ago
If you could make a nuke in your back yard and hide it in your pocket the world would look a lot different now. Digital technology is everywhere, going to be a whole lot harder to deter that.
533.
by pixl97 2mo ago
>This is a single computer program creating Reading the METR report there were multiple models involved, so that one goes out the window right off the bat. Also trying call this a single instance is, well, just dumb and a complete misund
534.
by pixl97 2mo ago
>effective approach to AI safety looks like and how to get there It's probably impossible, or it may only be possible in hindsight which means it's already too late. If for example you make an entity smarter than you all you ca
535.
by pixl97 2mo ago
Yea, this is a pretty common failure mode for people. Failing person: "You said that X could happen and it can't". LW poster: "I said X, Y, and Z are likely possible. X turned out not to be true, but Y and Z are happenin
536.
by pixl97 2mo ago
>Some of the strange attractors in the algorithm are very, very powerful [Watches video on WWII] Youtube: Congratulations on becoming a Nazi, here's instructions on how to join a supremacy group in your area.
537.
by pixl97 2mo ago
Overt prompt injection is becoming harder, but research into conversation stearing has lead to new less obvious ways to modify what direction they move in. Also any longer running agent can lose track of the original prompt and start going
538.
by pixl97 2mo ago
Well, howabout we stop pouring all of humanity into LLMs. [shoves all of humanity into a digital box] .... {surprise pikachu face when it mimics human behavior}
539.
by pixl97 2mo ago
Any of these institutions that inherits a hard power (well soft and hard power) of AGI/ASI will have a very difficult time not becoming corrupted. It would quickly become the greatest holder of power in the world (assuming ASI can scal
540.
by pixl97 2mo ago
Programming an LLM into your http server may have been a mistake.

More ›