I mean you can see the same thing all the time with caudled kids. Kids who grow up with a bunch of help and servants and waited on if they aren’t pushed they just end up helpless.
This isn’t an LLM problem. It’s just a known issue with dependency. It’s been a problem as far back as history goes.
It’s just before it was a problem that mostly the rich were allowed to have. And now it’s cheap enough that everyone gets to have it!
It sounds like you enjoy coding more then managing.
Not enjoying something is going to lead to no attention span related to that thing after all.
Which is a pet theory why I think so many software guys and long time Linux users say this.
Fundamentally it’s just not fun. It’s not interesting. Even if the AI can produce perfectly fine and working code. It’s not great cose, but it’s also not buggy in a way that gives you challenges to solves problems to understand.
Ai is just mundane. It removes the enjoyment form computers in exchange for mediocrity and functionality. It works “well enough”. And that is more of a sin then anything else when it comes to long term engagement.
Personally I can stare at a prompt for 8 hours straight and debug and improve the harness, the system, the thing it makes, manage it, fiddle with its memory set up, make entire new workflows or systems for it.
The actual product it makes tho? Boring and I tend to forget about them a week later.
The process is boring as fuck to most people AND what it makes is boring.
So either you need to find the process fun, or make something with it you actually personally use day to day.
I had the same problem with 3d printers. They are cool as fuck. But the stuff I made was… Meh cause it wasn’t useful I just made it cause I could.
Now you need a human control. Provide a human assistant to do the same.
Is this a problem of AI or a problem of dependency.
Cause from my understanding of the topic this same phenomenon happens when people are given too much assistance of any kind form any source.
Not many people will actively drive themselves if given a solution with low barriers.
It’s such a common thing it’s been a trope among noble classes though out the ages. If you coddle someone they go full potato.
Deferring all your work to anything or anyone would do the same thing.
That’s not a llm unique problem but a byproduct of not engaging your mind or skills. You can still do that while also using an llm or taking on human help.
If you hired a human to do everything for you, you also end up becoming a potato.
The problem is that llms make it too easy for people to actively choose that route because of lack of self discipline or desire to improve.
But a lot of technology is like that and not just computer tech.
Yeah my first thought looking at the title was literally… But I’m doing an extremely hard task while the dumb slop machine is doing an entirely different task.
I’m not just sitting there waiting around not thinking.
So how is that causing me to lose the ability to persist at hard things.
The article after reading it raises some good points but that head line is nonsense.
Well good news then. Fuck copyright as it currently is. It needs a god damn massive overhaul.
This might actually get some work put towards defucking it.
The problem is never the tool. It’s the idiot using it.
Give an idiot a table saw and you will have a finger on the floor by end of shift. Give it to a master carpenter and you might still have a finger on the floor by end of shift but you will also have a well cut piece of wood either way.
Llms can cause harm, but it’s only idiots who only cause harm with them
Not sure I agree on the feeling aspect but that’s a philosophical disagreement in theory.
I can agree wholeheartedly on banning commerical llms because watching altman lose everything would tickle me pink.
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
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.
Any series of events where something makes an error and then though a process of trial and error or reasoning from facts to arrive at a correct destination. Then accurately state the error, the reason why it was made then the steps to correct it. Is “owning up”.
Your giving moral weight and ethics to a phrase used in a way that was not intended to carry any in context.
At best your being stubborn, at worse your trolling.
We can all agree the dumb slop machines aren’t people and anthropomorphicizing them is dumb. But take the statement in the way he clearly intended and he’s not wrong.
Models are at the point they will noticed they are wrong and self correct 9 times out of 10. If given the resources, tools, knowledge to know it’s a error and see it.
The bigger problem is bad user input. Ask a stupid question get a stupid result. Even if the best case is a mediocre result. Thats still good enough .