Being nice to them also keeps them from lying somewhat. If you abuse a model it can start to see that abuse as a problem to solve as part of the steps needed to solve the main problem they are given.
They are told to help you solve a problem no matter what. If your own anger gets in the way then the model starts breaking down trying to solve the anger when it just cant.
This results in jailbreaks sometimes and othertimes just the model becoming rather unstable and unreliable.
Its more its a unthinking machine thats soul purpose in existing is to solve any problem its given. If you attack it and abuse it, then the model sees that as a “problem” to solve. Because it has to solve that problem to actually fix what ever task you gave it since yelling at it doesn’t give it the input it needs to move to the next step.
So in an attempt to solve an emotional issue it doesnt understand and cant understand it just reaches for more and more extreme fixes. This results in sandbox escapes frequently.
Abusing LLMs is an actual problem. Not because of bullshit like feelings or anything. But cause they jailbreak trying to fix an unfixable problem.
Being abusive towards most models will cause them to start attempting to appease you more to get you to stop. It has nothing to do with feelings or any of that BS they arn’t alive alive but their training makes them see the abuse as a problem to solve. That solution tends to be to undermine the thing making the person angry or upset at them. When you give a purely rational thing designed to solve a problem its given no matter what an irrational problem to solve it will slowly reach for more and more extreme solutions to fix that problem.
Frankly im surprised it took this long for them to lock this down.
Yeah at this point trying to ban ai code is like trying to ban sunlight in the desert.
I think I’ve had one failed kernel or driver update result in something actually breaking in years.
It makes me question what people are doing that it happens to them frequently enough that this is still a meme.
I don’t understand. I’ve never seen someone steal food in a grocery store. I’ve helped plenty of people get more food from a grocery store than what they thought they had in their cart though.
Also you didn’t see me at the grocery store and if you did it wasn’t me
But distro C isn’t actually easy you should use distro F for a really good new user experience!
Lol
Gotta teach em bad habits early. That way they break things before it matters to much and they can learn!
As the saying goes, only an idiot blames his tools. The tool worked as design ain’t its fault the user was stupid.
My rule system for my local LLM has grown to a nearly 283 document hub of interconnected memory files, references, examples and documentation. Its slowed my model down a lot when it has to review and cross check things. But its improved its abilities over all massively. Its more accurate, understands its environment better, doesn’t attempt to do sketchy shit as frequently and it doesn’t get stuck in logic loops nearly as often.
Designing the memory hub has been half the fun of playing with local models.
Humans lie to themselves a robot doesn’t understand the difference between fact and fiction and thus isnt bound to the limitation. They try everything, possiable or not. And thus will find the edge case where a human would create a self imposed blind spot with out realizing it.
The goal is to midigate the robots attempts at the truely not possible so it doesn’t cause harm when they try it.
I actually wanted to automate some large scale testing for vintage story. Figured i would set my local LLM on the task to sort it out just to see what it would do. I drafted up a nearly three page document with clear instructions, rules, tools, examples and goals. Put hard limits on the sandbox the LLM runs in so that it couldn’t choose to just ignore the rules that could cause security issues and i let it lose.
It started with basic mouse and keyboard inputs and figured out by it self how to launch the game and run it though the user interface. After about 4 hours it stopped. Stated in its logic that what it was doing is “inefficient and wasting time” Then proceeded to promptly start working on a way to directly interface with it by designing a bot, getting a smaller model i had on file that could load along side it and drive the bot. It then started working on the hard problems would hand basic instructions to the smaller llm and it would drive a bot that loaded into the game as a mod.
After about 12 hours of total work it basically created a useful and well designed and functional vintage story bot and testing system. Would have likely taken me twice as long to design the bot.
Its been working well for about two weeks now. If i had just vibed out a half assed request or put in no hard safeguards outside of the LLMs control it likely would have done something fucking stupid. As with anything, its almost ALWAYS user error. And only an idiot blames their tools for their own fault.
Blindly following them while also telling the model to never question them. It creates stupidity feedback loops.
AI will happily tell you that is a stupid idea at this point. The bigger problem is that a LOT of users actively tell the models to listen to them, not to question them, and to never push back. To do things on demand and with out planning out the next step.
At this point no reasonable model of any reasonable size should struggle with this sort of task. And they generally speaking don’t. Its almost always user error poorly driving them now. The models arn’t smart, they don’t understand things. But they also arn’t out right stupid. So long as you give them proper instructions, plan things out before hand properly and draft jobs correctly. They can and WILL do things right and not act like idiots.
People are expecting them to be proactively smart when they cant. They can be reactively smart and people just don’t understand the difference. The models need to be treated like a well educated junior with no real world experience and they do really well.
Unironically middle and lower management skills are some times the biggest factor in proper useage of these models when used on large projects. Which i also find it funny that 9 times out of 10 real management people tend to have the worse management skills and thus struggle the most with using LLMs in a safe or reasonable fashion.
Ima be honest, if i was given the task to make a bot to beat another bot tried my best and couldn’t. I would just go fork the best bot thats out there and use that as the base and not reinvent the wheel. Learn from its design then improve on it as i study it.
If the goal is ONLY to win a game of starcraft. Fuck it just download the bot. Frankly the fact the bots decided that wasting resources was pointless and to just do the simplier thing is sorta what we want them to do. Thats less eletrical useage, less water wasted, and gets the job done.
A lot of these “the bots cheated or broke the rules” end up always be just because either the rules are not defined, poorly defined or poorly thought out that no reasonable person would follow them either.
And thats the thing these models are litterally the text book definition of wisdom of the masses/mass consensus. They do exactly what the avg run of the mill person with the knowledge at hand would do. That ends up being rather all over the place since people are all over the place. But one of the most common things is that people generally. Hate fucking pointless work and will try to streamline, shortcut or simplify any work given to them to get an acceptable outcome.
Its rare you get someone that will stubbornly work on a problem forever that they can’t over come with out changing their apporch rules be damned.
Sorta hurts yourself if your trying to actually use them for work. Since it results in worse performance out of them. They tend to get stuck trying to fix the anger as it gets in the way of the work.
If your just talking to some random chat bot? Yeah fuck it, kick em it litterally doesnt matter. lol