News for hackers. Stuff that matters.
Re:Douglas Hofstadter changes his mind on Deep Learning and ... by arisAlexis on Monday July 03, 2023 @02:27PM (#36576995)
Loling something that serious with several top scientists as you said warning is at least very idiotic regardless where the truth lies. Let's start from that.
Re:Douglas Hofstadter changes his mind on Deep Learning and ... by ke88y on Monday July 03, 2023 @02:31PM (#36577055)
I'm laughing at the study methodology, not the underlying topic.

Please don't call me idiotic.

I'd wager I have spent at least 10,000 hours more than you working on (real) safety analyses for (real) AI systems, so refraining from insinuating I'm dismissive would also be nice. But refraining from "idiotic" seems like a minimum baseline.

Re:Douglas Hofstadter changes his mind on Deep Learning and ... by arisAlexis on Monday July 03, 2023 @02:36PM (#36577115)
Again, I repeat everyone that is loling at x-risk an idiot and that includes many high profile people with huge egos and counter culture biases. (Hello @pmarca). There is a big movement to call ai doomers many names and generally make fun and dismiss the risk. It is exactly like people laughing at nuclear risks saying its not possible or not a thing even when Einstein and Oppenheimer were warning us. If you belong in this group is up to you.
Re:Douglas Hofstadter changes his mind on Deep Learning and ... by ke88y on Monday July 03, 2023 @02:39PM (#36577152)
> It is exactly like people laughing at nuclear risks saying its not possible or not a thing even when Einstein and Oppenheimer were warning us. If you belong in this group is up to you.

In his letter about the potential for an atomic bomb, Einstein's recommendation to the Defense Department was to start work on the bomb ASAP. It was a warning about the adversary getting the technology first, and did not call for abstaining or for work on nuclear safety research. Those came much later, after the risk was real rather than hypothetical.

Re:Douglas Hofstadter changes his mind on Deep Learning and ... by vhlhvjcov on Monday July 03, 2023 @03:35PM (#36577784)
For those of us who don’t think you are an idiot (I don’t), could you maybe give us your insights and a history lesson?

I am most intrigued, particularly regarding Oppenheimer and Einstein.

Re:Douglas Hofstadter changes his mind on Deep Learning and ... by ke88y on Monday July 03, 2023 @03:57PM (#36577995)
The tldr is that both of them urged Roosevelt to develop the weapon, and only later when the destructive potential of the bomb was obvious expressed regret. Einstein's 1938 letter to Roosevelt was the first step toward the Manhattan project. See https://www.osti.gov/opennet/manhattan-project-history/Resou... if you want to read more.

So it's weird to say Einstein "warned us" about the x-risk of nuclear weapons prior to their development when his letter to Roosevelt was begging for more money to speed up the development of nukes.

I think the entire saga is mostly an unrelated red herring -- AI is nothing like nuclear bombs and reasoning by analogy in that way is sloppy at best.

Mostly? It's just kind of funny that x-risk people point to Einstein and Oppenheimer as positive examples, since they both did literally the exact opposite of "warn the public and don't develop". The irony makes you chuckle if you know the history.

Particularly given the weird fetish for IQ in the portion of the x-risk community that overlaps with the rationalist community, it's also really funny to point out that what they should be saying is actually something like "don't be like those high-IQ fools Einstein and Oppenheimer! They are terrible examples!" ;)

Re:Douglas Hofstadter changes his mind on Deep Learning and ... by arisAlexis on Tuesday July 04, 2023 @02:25AM (#36582928)
Not at all. My point was that while Einstein said (warned) that nukes are possible and while they were being built in Los Alamos, the lower key physicists,academics and apparently people like you were saying that nuclear explosion is impossible. It was the quite laughable. And it's exactly the situation now, Agis are being built and laymen are in denial.
Re:Douglas Hofstadter changes his mind on Deep Learning and ... by arisAlexis on Tuesday July 04, 2023 @02:26AM (#36582936)
> you're willfully misconstruing what I am saying and then calling me an idiot

No I am not, you put yourself in the group that makes x-risk sound ridiculous on purpose. If you do that yes. Can you make yourself a third person listening to someone saying: "important people said there is a chance everyone dies but I am laughing at them"? How psychopathic is this?

> If I believed you were operating in good faith

No, you are right. I have a back agenda, it's a dark one. It is about not everyone dying. Call me a backstabbing lying scam.

> why my life experience makes me suspicious of the personalities involved in pushing AI x-risk

Mistake no1: "pushing x-risk" makes no sense. Being afraid that we will die makes a lot of sense, humans are built this way. Some take action some laugh etc. Mistake no2: Hinton, Bengio, Ilya Sutsekever that is the guys that built the technology you spent thousands of hours researching as you said. If you are suspicious of these guys, just get out of this field, it's not for you clearly. That would be insane, a physicist being suspicious of Feynman or Bohr.