Comments by "" (@obsidianjane4413) on "AI regulation is coming..." video.

@samueleu6362 What could possibly go wrong...
27
@TheNewton "ML takes a gigchads of compute power AND data" Currently. Until somone/thing develops the better mousetrap. That is what has these guys scared. If something develops a distributed architecture and escapes to the millions of processors of the Internet it has infinite growth potential and will be effectively unstopable.
13
@MCLVideo Manufacturing a knife that could kill millions of people is probably not a good idea.
8
@MCLVideo Perhaps you should educate yourself a bit more on the subject than just one tongue in cheek video? Real life is harsher than video games and you don't get to restart.
7
Thats like assuming that giving everyone an atomic bomb will prevent nuclear war....
5
@dtibor5903 If they are developing something that could be dangerous if release into the wild, shouldn't they?
5
@facundosterzerforino1704 Not really. Currently for AI to be potentially dangerous, requires scale and computing power. You training an AI to do trivial tasks like automate or expert at something probably won't be covered.
4
@jwonz2054 "The fear mongering of AI replacing workers is currently overhyped." Right up to the moment when its not.
4
@rumfordc That is not true. Please go educate yourself.
4
@sometard8156 Potentially without humans if the AI decides the world would be better off without us? Someone tried that last century. It didn't turn out well for anyone.
3
@marcelh7864 Then you misunderstood my post. Government(s) want to make sure that IT is that person who controls that first AGI and bends it to its bidding. They and the rich and powerful will want to maintain the status quo, which for better and worse, is probably the best option for the public than the unknowable outcome of a wild singularity.
3
No if they knew that had happened already, all of these guys would be hiding in their underground lairs and private islands.
3
Only when the state is sure of its own monopoly on power. AI potentially challenges that. So it unintentionally serves the public interest. Hopefully.
2
@marcelh7864 NP. It was tucked in around "don't take ma freedums" gibberish.
2
@sairos4057 When was the last time you died of food poisoning?
2
@sairos4057 lol. I'm not gonna tell you to go educate yourself because you might be from a place that doesn't effectively regulate much. But goo educate yourself.
2
@rumfordc lol. You think AI is just autocorrect? Bless your heart.
2
@rumfordc Wrong again.
2
We'll hopefully have a lot of time to work on that while hanging out in our machine curated primate reservation or hiding in the rubble of our cities...
1
Ah no. Government will still get its cut in whatever profits AI produces. And governments fund a lot of AI research in its own interests. What government doesn't want to happen is disrupting its status quo.
1
TLDR: We're all doomed.
1
@whatever2144 lol. "Potential enemies" (and friends, and your own government) are not going to publish their development in OS, they aren't publishing anything. What you see in public is only what those want you to see. Regulation isn't going to change that. All it will change is maybe slow down and inhibit non-state actors or mid-capability organization (companies and universities) from potentially creating something unethical or dangerous. Ironically, you would need to regulate (pass a law) that would force companies to disclose their tech.
1
@dtibor5903 And you even edited this ignorant gibberish?
1
@Dmitrii-q6p Well if you dictate alignment then you are restricting companies. What happens when it learns to change how it teaches itself? When someone one gives it access to the raw, unfiltered Internet? What happens when it learns to manipulate physical machines and IoT devices? What happens when all of the forks become dangerous and all of the news is fake? Hmmmm?
1
@facundosterzerforino1704 I don't expect good faith from politicians any more than I expect asshats to not do stupid so that the rest of us can't have nice things.
1
@maximiliandmc Do you understand that ChatGPT isn't going to do anything but maybe scam some people out of money, but the MI underlying it has wide application and the ability to develop independently? A GPI isn't going to talk to us while its doing whatever it decides to do. Perhaps it is YOU who is not as intelligent as you think you are?
1
@ThePallidor You really didn't think that one thru or you are not aware that laws that courts judge with are created by legislators and regulators. If AI is unregulated, as it is now, the courts can't do anything besides help clean up the damage with civil suits. They have zero ability to prevent it in the first place.
1
@rumfordc AI is machine learning. Its software that teaches itself. Auto correct and even ChatGPT is just one relatively trivial (but lucrative!) application of the technology. But it is being applied to many subjects and fields. It is learning things that humans are not programming it to do and we don't know why.
1
@rumfordc I don't argue with stupid people. Feel free to stop demonstrating you don't know what you are writing about.
1