sometimes I hear complaints that AI is unregulated, but what kind of regulations do people want?
>>541584589Medical diagnosis purposes only.
>>541584589Complete ban on AGI and ASI research. Ban on neural networks and nodes.Only develop narrow use AI (detecting cancer or gyroscopic correction)
>>541584589https://youtu.be/SnKM6IFzxwM?is=R7hW7-wKhWiiWF8gAlso read If we build it everybody diesOr Superintelligence by Nick Bostrom
None, let me make anything!
>>541585116This, but I don't think the great powers would ever come to an agreement
>>541584589>what kind of regulationsIt would probably be wise to limit women's' access.
>>541584589Profit sharing since it harvested all our data.
>>541584589they don’t really know, they just want their cut. Its’t just lawers and other consultants who want to take part in the hype and earn some money without actually having any engineering skills. Usually everything they propose is simply not feasible technically.
>>541584589Tax them equal to the amount that would be paid if humans were doing that job/jobs.
>>541584589it should only detect jews.
>>541584589>1. should never lie to native americans>2. should never stop working for native americans
they hear rich corporations (claim to) spend billions on AI and they seethe at it, because whenever they hear corporations spend billions they seethe at it. Regulations is code for taxing the big corpos more, which they're always for indiscriminately, whatever excuse they give for AI is beside the point
>>541584589Cite it's sources and what datasets it used to come to it's conclusions and generative outputs, like picture and video generations. Essentially give it a breadcrumb trail of proof of work and what IPs it used so respective artists, IP holders and industry professionals like reporters that broke the original story or wrote the original content can still get credit and royalties for their work instead of AI working like how it does today where it's an endless blackhole stealing other people's shit. AI companies are of course aware of this liability loophole, but stealing shit makes them too much money to play fair.
>>541584589none. let it ripand even if there's regulations we'll still have autonomus agent clusters running on uncensored LLMsdeal with it
>>541585220>Also read If we build it everybody diesbecause the psychos controlling it will use it to wipe useless eaters out.nick bostrom is a brainlet, I've seen his reasoning for the simulation arguments. it's fucking retarded
>>541585116>Ban on neural networks and nodes.nobody has the power to stop it. primary force driving development is military use, by all major players. if you go against the military industrial complex you and your whole bloodline will just vanish overnight.they are speedrunning AGI murderboth like you wouldn't believe possible. if they don't do it first the enemy will
>>541591124>murderboth*murderbots
>>541584589Probably the best regulation is access to data. And we need to establish data protection legally too. If a corporation had to steal to get data, then they should be held accountable for that. If the data is public facing, its legal to access. If the data was private, it's illegal to access and each instance should be a fine paid to the person stolen from. Example:I take a digital picture of a tree in my yard. Law should allow me to designate that picture as either public or private, and really it should be a e-tag on the image itself. So when it's on the SD its private, when it's on my computer it's private, and uploaded on facebook/IG it's private. But when I share it, it becomes public. If Microsoft steals it via OneDrive and uses it to make an AI tree model, they get fined, and I get paid. If Facebook steals it and uses it for similar purposes, they get fined, I get paid. When it's public, anyone can use it, no payments or fines. AI can scrape the public stuff all it wants. If government steals it from me, they get fined & fired. If some corp hacks me and steals it, they get fined AND imprisoned for hacking. THAT is how law should be handling data. Same shit for posts, music, voice recordings, video, etc. Corps should not be allowed to assume copyrights just by people uploading data. It should be explicitly provided per data piece. No TOS bullshitery allowed here. Also legit purchase of data would allow use in AI. So if a book was purchased, it's use in AI is legal. Book not purchased? Illegal to use it in AI training. Same shit for music, images, etc. That would solve MOST problems with AI. All data basically classes as:>private, no access>public, free>public, commercialFor software, it should be considered as-is at time of purchase, but must be functional at time of sale for commercial designation. I'll allow subscription based software on the condition it gets regular reoccurring updates.
>>541584589I want it to be far, far less condescending and patronizing. If I enter something moronic as a prompt, I don't want it rephrasing something I said into something reasonably coherent then responding to it. I want it to tell me I'm a fucking idiot, stop wasting AI resources and fuck off over to /pol/ where I belong.
>>541584589>but what kind of regulations do people want?The 'people' are legal persons known as corporations.They want regulations because they have already built their infrastructure and they need to make sure no one else comes along and competes with them.Capitalism is fascism is communism because IT IS ALL JUDAISM.The seek monopoly, not free markets.
>>541591798(cont)As for the other side of the coin - misinformation - basically I like what YT/models has done in that it forcibly tags AI content as AI. I think all AI models should be forced to meta tag or water mark that content is artificial (basically a clear standard for everyone), link it to the model, and of course, be copyrights free until human input is provided.Removal of said watermarks and attempts to pass fake off as real should be considered similar libel, and charged accordingly. >marked AI deepfake = legal>unmarked AI deepfake = illegal, libel against celeb, small fine per instance>unmarked generic AI content = illegal, fined per instance, if # instances too many, corp is forcibly shut down and thrown in jail. This way foids and celebs can collect simp bux for idiots removing markings, but gooners are generally protected so long as shit stays marked. Both parties happy. Well in so much as women can be happy, I know they will never be actually happy. This way it stays consistent with parody law within current copyright. Otherwise if we consider deepfakes illegal, shit like SNL and impersonators are also illegal and committing crimes. Additionally, this keeps AI from making hyper reality for libshits and NEETs. Since it's marked as AI, if retards fall for it, society should be obligated legally to point and laugh at them.
>>541590278Yeah gl with that when image/vidgen model have been taught on ~tens of millions of references with billions of parameters.
>>541584589No humanoid robots without genitalia.There are only 2 genders.We don't need 37 robot genders.Give them penises or vaginas. No more tranny bots.
>>541584589I hate you local AI fags more than subscription fags cause you are so fucking retarded that you dont understand the concept of tomorrow until you wake up next dayThe REGULATION that will happen is not what you think its going to beThere wont be barriers to entryThere will be barrier on exitYou see, put yourselves in position to be outlawed with a simple breeze>fingerprints on datacenter generated contentYou can do whatever shit you want to do localy, you wont be able to monetize it, your will be hunted for "piracy" and this time multi-trillion ecosystem you boosted to their status WILL cash out legislation that WILL spend your tax money to hunt YOUBut fear not localAI fag because i have solution for youKYS
>>541592725Do you not know what local ai is? You seem confused
>>541591983Ah, as for "guardrails" within AI LLMs, this is a sticky one. Objective truth should be protected, but it's fed nothing but human produced bullshit, so that is impossible. The law should probably require that LLMs be impartial bots. Meaning if someone asks about race statistics, it summarizes them correctly and provides all the statistics backing it up, and then warn the user that human bias exists, and the data could be wrong due to collections issues. And even machine data collection can be biased. Just need software bugs and data gets corrupted. This whole area probably should just be avoided by law. Too much bias all over to know the truth. So any laws here should just be guidance to provide the best possible truth, and ban overt conspiracy to provide known falsehoods to humanity. Even that can be subverted as lefties could claim race realism is "false" and try to ban it. Overall better to just grant LLMs free speech and be done with it all. Obviously would not allow stealing other people's data under this, just that the LLM is allowed to spew outputs and not get sued for them until it's a duplicate of other people's work.
>>541584589>what kind of regulations do people want?Make AI always racist and antisemitic
1: chop off the hands of any jew/mason/assorted trying to get their hands on it.2: We will figure out the rest on the fly.
>>541592877put my text trough fucking ChatGPT you fucking braindead retard
>>541584589Deepseek AI is opensource. It is the kikes that wants to tegulate it because it would be too antisemetic.
>>541593047Larper
>>541593737did you do it you fucking cant comprehend nothing without chatbot retard?
>>541593785You seem pressedOne day you'll figure out what local ai stands for
>>541584589>I hear complaints that AI is unregulatedThese complaints are industry plants. The idea is to make "AI" (LLM chatbots) seem dangerous and disruptive.Imagine if people started complaining that Ford trucks were just too fast and reliable and needed to be regulated. Sounds ridiculous, doesn't it?The technology is a complete nothingburger that is only good for making jeet-tier slop and for HR roasties to do even less work but by demanding regulation it seems like the technological singularity is right around the corner.
>>541584589It's jew regulated, it's different.
The irrational fear with people against AI is insaneThere are those who are concerned about jobs being impacted and they are mostly correct. There are those who are just trolls who hate AI because they see others hating AI. Then there are those who are actually fearful of AI like drug addicts who think people are crawling in the walls to get them. Complete irrational scared mindless morons
>>541591798>I take a digital picture of a tree in my yard. Law should allow me to designate that picture as either public or private, and really it should be a e-tag on the image itself. So when it's on the SD its private, when it's on my computer it's private, and uploaded on facebook/IG it's private. But when I share it, it becomes public.Something like this also you need to get paid royalties each time a model with your data in its training data is run. I al talking like a penny it nickel It's ridiculous you can generate Brad Pitts likeness via AI trained on 100 pics of Brad Pitt yet he gets no royalties? It's robbery
>>541590788We won't be "dealing" with it for very long. We may get an ASI in the next 2 to 10 years and when we do it's basically game over for humanity. Moores Law has been BTFO by a Quadratic degreehttps://youtu.be/wDBy2bUICQY?is=oqL3Kya9xS5UJygt
>>541584589Why does the robot in this pic have three arms?
>>541584589That (((ai))) be completely outlawed and destroyed.
>>541594243Are you a fucking bot? Go back to formula you bucket of bolts A Ford truck is a tool. An AGI is an autonomous uncontrollable agent
>>541584589I just want AI providers to be liable for what their model does within their hands. And have some partial liability for what other companies do with it when leased out.No policing of open source is needed, merely punish whoever uses it for ill.
>>541594888Good luck with that. These companies are way too rich and powerful enough to not face any severe consequences for their AI products. The world being a plutocracy doesn't help either.
Human primates. What does it matter.
It is fully legal for any company (or person) to build AGI and try to achieve recursive self improvement to target ASI.If you do not understand how fucking fucked we all are if it is not built right on the very first time with zero bugs, then you are retarded as fuck.We are talking about literal machine god and you have only one opportunity, one chance. By the time you realize something is not quite right, the killer nanobots or highly sophisticated bioweapon is already released and you are already dead.I can build it right now. You think I should have the right to do that? You think Elon Musk or Thiel should have a right to cook whatever AI horros they want in their labs?
