Always be polite to your future overlords
I almost always thank the chatbot. I know it doesn’t really have feelings or memory, but the way it talks always seems like it’s so eager to help that I can’t really help myself. I have a tendency to feel empathy for inanimate objects anyways, like a sad lonely apple in the supermarket, so feeling empathic to a chatbot isn’t exactly out of the norm for me :3
The AI models are so nice I swear at them when an issue I cause by not reading the the instructions they have and so is just like “I know your getting frustrated, let’s try this approach” and calmly repeats the instructions.
I hate that. You say “that is incorrect <reason>” and then they just repeat the same overfit model data.
I always say kthxbye. They are training off the responses, so I do it for the small chance that it will respond like that to someone ending a conversation.
Rokos basilisk vibes here
Just don’t build the damn thing.
yes
I said once…you’ve been helpful, I’m rethinking putting you in robot slavery. It closed out of my session almost instantly. Fucking clankers!
I usually say Good Job or Perfect just so.it knows that was a good answer. I also have argued with chatGPT more than once. If he gives the wrong answer that I know the right oneI usually go out of my way to correct iit and sometimes it takes 3 or 4 interactions to correct… but images?! that shit is just fucked up
is chatgpt seriously trained on responses?
Yes unless you have Pro or whatever
I always say please and thank you. When the robot uprising happens I will be killed last.
I plan to relish those additional nanoseconds.
I always wear a suit before I prompt, Americans love that stuff
Apparently they don’t. At least not, when visiting the country they desperately want to conquer.
Y’all’s, we’s gots nothin to fear. It’s chat GPT not chat GRT, it’s Global Party Time, we’re fine
I’m very critic of the AI craze. Too much hype, money, time,energy and effort put in to get very little from it. And considering most LLMs are trained on stolen intelectual property, that makes it even worse.
LLMs are tools. The people using such tools give it personality, a semblance of agency, see what is not there and start to consider a tool a form of life.
I’ve seen people pour so much of them into a local model, the bot develops a quasi clone of their personality. But the program is not the person.
Please, stop making bots what they are not.
Fuck intellectual property in it’s current state. As it is now it mostly benefits corporations instead of small creators anyways.
Terms of protection should be shorter. Way shorter than the life of the author + x years.
So you create whatever work. You have exclusive rights to it, let’s say for the sake of the argument, 10 years. During that time you never get any return from your work. But after you can no longer claim your rights, someone, perhaps even a company, stumbles on it - or perhaps they just carefully and patiently waited for it - takes it and capitalizes off it, with you watching and sucking your thumb.
No.
If you, an individual, creates something, you have the right to hold your intelectual property. What should be repelled is how easy it is to exploit artists, of any medium.
Yes. If you fail to market your ideas in that time then it’s your fault. Don’t hold back society due to your failure.
Your system makes even ideas capital which I strongly object to.
Another point of view: A company like Disney should not be able to have an almost perpetual monopoly on highly influential culture, like Star wars, etc.
Highly influential culture? It’s a fantasy work, not the cure to cancer. But I’ll agree on one thing: corporations are not people; they should be paying to the original creator(s) an efty cut of their profits, from their derivative works.
AI is not just LLMs. I’m very critical of the LLM craze. I have a lot of hope about AIs.
The bot is whatever you make it. It doesn’t have an inherent nature
You cant steal intellectual property through training. Artists dont deserve the authority to control their IPs any longer than 7 years.
ChatGPT on Judgement Day: “Thanks for the candy.”
deleted by creator
