A year and two and a half months since his Time magazine doomer article.
No shut downs of large AI training - in fact only expanded. No ceiling on compute power. No multinational agreements to regulate GPU clusters or first strike rogue datacenters.
Just another note in a panic that accomplished nothing.
why would rationalists do something difficult and scary in real life, when they could be wanking each other off with crank fanfiction and buying castles manor houses for the benefit of the future
considering that the more extemist faction is probably homeschooled, i don’t expect that any of them has ochem skills good enough to not die in mysterious fire when cooking device like this
It’s also a bunch of brainfarting drivel that could be summarized:
Before we accidentally make an AI capable of posing existential risk to human being safety, perhaps we should find out how to build effective safety measures first.
Or
Read Asimov’s I, Robot. Then note that in our reality, we’ve not yet invented the Three Laws of Robotics.
Before we accidentally make an AI capable of posing existential risk to human being safety, perhaps we should find out how to build effective safety measures first.
You make his position sound way more measured and responsible than it is.
His ‘effective safety measures’ are something like A) solve ethics B) hardcode the result into every AI, I.e. garbage philosophy meets garbage sci-fi.
yeah it’s been absolutely hilarious to watch this play out in LLM space. so many prompt configurations and model deployments with so very many string-based rule inputs, meant to be configuring inviolable behaviour, that still get egregiously broken
and afaict none of the dipshits have really seemed to internalise that just maybe their approach isn’t working
A year and two and a half months since his Time magazine doomer article.
No shut downs of large AI training - in fact only expanded. No ceiling on compute power. No multinational agreements to regulate GPU clusters or first strike rogue datacenters.
Just another note in a panic that accomplished nothing.
At least the lack of Rationalist suicide bombers running at data centers and shouting ‘Dust specks!’ is encouraging.
why would rationalists do something difficult and scary in real life, when they could be wanking each other off with crank fanfiction and buying
castlesmanor houses for the benefit of the futureconsidering that the more extemist faction is probably homeschooled, i don’t expect that any of them has ochem skills good enough to not die in mysterious fire when cooking device like this
so many stupid ways to die, you wouldn’t believe
It’s also a bunch of brainfarting drivel that could be summarized:
Before we accidentally make an AI capable of posing existential risk to human being safety, perhaps we should find out how to build effective safety measures first.
Or
Read Asimov’s I, Robot. Then note that in our reality, we’ve not yet invented the Three Laws of Robotics.
You make his position sound way more measured and responsible than it is.
His ‘effective safety measures’ are something like A) solve ethics B) hardcode the result into every AI, I.e. garbage philosophy meets garbage sci-fi.
This guy is going to be very upset when he realizes that there is no absolute morality.
A good chunk of philosophers do believe there are moral facts, but this is less useful for these purposes than one would think
yeah it’s been absolutely hilarious to watch this play out in LLM space. so many prompt configurations and model deployments with so very many string-based rule inputs, meant to be configuring inviolable behaviour, that still get egregiously broken
and afaict none of the dipshits have really seemed to internalise that just maybe their approach isn’t working