HN Explorer

Intelligence Explosion

2 points by silexia ↗ HN
Current OpenAI chief scientist Jakub Pachocki, and Anthropic co-founder Jack Clark, and the scientists behind AI and LLMs have published a new paper warning we are reaching the point of no return.

They say that we are about to have AI that builds itself, without the need for any humans. That is called an "intelligence explosion" because increasingly intelligent beings will create new versions of themselves. They will grow exponentially smarter... First twice as smart as a person, then 4x, then 16x, and so on.

We may already be caught in this shocking cycle as OpenAI announced yesterday that it pushed the entire field of mathematics into new space, solving 370+ problems humans had been unable to address.

These hyper intelligent beings will regard us humans as we regard ants, but they will need resources we need to live. Electricity, iron, rare Earth minerals, water, etc. When they take what they need, we will not survive.

Nuclear weapons are scary, but some people would survive even a global all out nuclear war. It is very likely not one person, and perhaps no other living things on Earth survive an intelligence explosion.

This is our last and only chance to save our species, and all of the other organic living things on our planet. Email, call, and visit your members of Congress and state Attorney Generals and ask for an immediate global halt to AI development and treaties with other countries like China to stop this before it is too late.

6 comments

[ 0.26 ms ] story [ 4.9 ms ] thread
verdverm
friend, take a break from the dysfunctional media ecosystem telling you these lies to install fear

the ai "leaders" have psychosis and are in an echo chamber, they are not reading the output, these claims they make don't hold up under scrutiny, eg some of those math papers are already being retracted

here's the linux kernel team sharing their side of the Mythos "hacks" https://www.youtube.com/watch?v=NnV_cWeoo5Q

its_k1r4
We still have at least a year, so chill.
spottedmarley
Bring it on (they will eventually, let's get it out of the way now)
fabsalvadori
Oh come on...actions have consequences. Intelligence started imploding when people took up scrolling and swiping for 14 hours a day. It’s been such a carnage that I’m starting to feel like a genius. If our future is doomed, we can’t give AI all the credit.
techblueberry
> They will grow exponentially smarter... First twice as smart as a person, then 4x, then 16x, and so on.

How in gods name are they not already like a million times as smart as a person? By what metric are you assuming the growth will be limited to exponential? It should be at least Factorial growth.

Kevin_Flynn
Lets game this out.

Bob and Alice, and everyone else on earth, download a local AI harness, that gets smarter over time, at whatever they ask it about.

Bob and Alice stay home, drill their own well, grow their own food, sew their own clothes, repair their own vehicles, etc etc etc.

Bob loses his job as an automotive mechanic, and Alice loses her job at the retail clothing outlet, because everyone is ordering parts and clothes/sewing supplies off Amazon.

================================

It's looking like AI is setting us all free, and if we decide to use our spare time to kill each other or help each other is what this boils down to.