Here we go. Reporting requirements:
(i) any model that was trained using a quantity of computing power greater than 1026 integer or floating-point operations, or using primarily biological sequence data and using a quantity of computing power greater than 1023 integer or floating-point operations; and
(ii) any computing cluster that has a set of machines physically co-located in a single datacenter, transitively connected by data center networking of over 100 Gbit/s, and having a theoretical maximum computing capacity of 1020 integer or floating-point operations per second for training AI.
First they came for the privacy advocates, But I believed I had nothing to hide, so I didn't speak up. Then they came for the Bitcoiners, But I never invested in bitcoin, so I didn't speak up. Then they came for the AI developers, But I thought machines were too complex anyway, so I didn't speak up. Then they came for me, And by that time there was no one left to speak up.
Key passage: " (a) Artificial Intelligence must be safe and secure. Meeting this goal requires robust, reliable, repeatable, and standardized evaluations of AI systems, as well as policies, institutions, and, as appropriate, other mechanisms to test, understand, and mitigate risks from these systems before they are put to use." Emphasis mine. This is similar to the FDA regarding medicine. Big players end up paying the regulating agency to "evaluate" the proposed solution. Takes years and is not effective -- unless "effective" means protecting incumbents who can afford to lobby.
Heh you should probably format the exponents correctly because when I read "computing power greater than 1026 integer or floating-point operations" I did a double take.
Oh crap you're right! I copied those without doublechecking.
The absurdity of this is breathtaking.
Meh, think of this way. It fills multiple needs:
Everyone wins, right? Except for actual quality of product and value to society.
The Noble lie also fits. I mean I use that as a lens to view these things as well.
I don't disagree. What I am doing is taking them literally. What you are doing (and what I think as well) is thinking about the game theory behind it. You nailed it. What you described is why we can't have nice things.
It's like they think restricting AI in the United States is going to restrict AI.
A few years ago I started viewing the State and specifically the U.S. Federal government as a post-theism god. It has many of the trappings of a religion and one of the biggest ones is this idea that the word of the State can just magically make things so. Of course this isn't true and isn't new but many seem to sincerely believe. They are offended in the same way a religious person is offended when you speak ill of their god.
This perspective has really helped me understand why so many people get so worked up about politics and their political team. It truly is a religion and as such when you speak blasphemy you should be silenced.
That's insightful. A bunch of people and their opinions make sense from that lens.
A common secular/cynical view of religion is that it's an anxiolytic, ie "the power is concentrated somewhere and I'm in its good graces." People totally use governments that way.
You nailed it. This is a human issue. IMO the issues with religion are issues with humans. You remove "god" and people create a god and act the same way.
Not that there aren't issues with all religions and people that practice them. Its just that religion is an easy scapegoat. Its just not that simple.
I wonder if "Government" knows this, and uses it to its advantage? Or even intentionally cultivated it? 🧐
(Spoilers: Yes)
HAHAHAHAHAHAHA! Funny
Yeah, its quiet stupid. The hubris of these people seems to have no end.
Ban AI. Ban Bitcoin. Ban Meat. Ban Encryption. Ban Guns. Ban Truth. Ban Productivity. Ban All the Things. Ban Self.
The End.
wish they would just skip to the last part