Slowing Moore's Law: Why You Might Want To and How You Would Do It | Hacker News Reader

Slowing Moore's Law: Why You Might Want To and How You Would Do It

gwern.net

28 pointsgwern17 comments

"[Whole brain emulations] are frequently regarded as being potentially highly untrustworthy and dangerous, constituting an existential risk."

Really? Are they? An ~8000 word article based on this one sentence, with weasel words and no real rationale. Also, you seem to be confounding the computing power required to emulate a brain ("straightforward engineering") with the new imaging techniques that would need to be developed to scan and "upload."

http://intelligence.org/files/WBE-Superorgs.pdf

> Because such emulations could be freely copied and run at increased speeds, they might quickly outnumber humans and be capable of performing almost any task more cheaply. Standard economic models suggest this could produce tremendous economic growth, perhaps doubling the size of economies every few weeks or less, but also driving wages for most jobs below human subsistence level (Hanson, forthcoming). Many have suggested that such rapidly replicating and evolving minds could cause human extinction if not carefully controlled (Bostrom 2002; Yudkowsky 2008; Posner 2004; Friedman 2008; McAuliﬀe 2001; Joy 2000; Moravec 1999).

So, we're the new Neanderthals?

While I'm somewhat convinced that human-equivalent or greater machine intelligence will likely happen within the next hundred years (or less), good luck convincing a majority of humans of that.

> An ~8000 word article based on this one sentence, with weasel words and no real rationale.

A more accurate summary would read "a 7900 word article on the little-known economics and vulnerabilities of modern chip fabs, with ~100 words explaining how the author came to be interested in the topic". You could replace the WBE bits with "I am an investor in Intel and was reading Taleb's _The Black Swan_ and began to wonder what black swans might affect my stock holdings", and the essay wouldn't be that different.

Fair enough. I guess I was more interested in the "why" than the "how" from the title. And given how thoroughly researched and annotated the "how" section was, I was surprised so little consideration was given to why in the world you'd want to do this in the first place.

> And given how thoroughly researched and annotated the "how" section was, I was surprised so little consideration was given to why in the world you'd want to do this in the first place.

Yes, well... If I may quote Nietzsche: "Lessing, the most honest of theoretical men, dared to say that he took greater delight in the quest for truth than in the truth itself."

I care very little for the 'why', but once the thought has arisen - 'how vulnerable are chip fabs, anyway?' - then I am completely fascinated with the 'how'.

Wat.

People take their singularity BS way too seriously.

Yeah, the background part of this article is great but the idea that anyone (especially governments) would voluntarily take a noticeable bite out of their economy to avoid the very small risk of the evil singularity sounds like crazy talk. (It just occurred to me that a comparison with climate change may be possible.)

> the idea that anyone (especially governments) would voluntarily take a noticeable bite out of their economy to avoid the very small risk of the evil singularity sounds like crazy talk.

Absolutely. And I say as much, at length, in the OP.

> (It just occurred to me that a comparison with climate change may be possible.)

It is not just possible, but I actually make that comparison in the OP. (See point 4 in the China section.)

Sorry, I didn't devote the time to read your article carefully and by the time I reached the end I had forgotten the beginning. On second reading I find this part a little odd: "...we would need to make a very solid case indeed - a case I cannot make nor will make here. Let’s assume that the case has been made and examine an easier question, how feasible such strategies are at accomplishing the goal at all." It's almost like you're sabotaging your own argument by saying that it probably doesn't matter but here are thousands of words on the topic anyway. (I realize this has already been discussed in the other subthread.)

I'm concerned that the singularitians (sp?) have perhaps let their knowledge of computing and economics go to their head, thinking that they understand philosophy and psychology just as well. But on the technical side, there's no reason to expect they won't be right eventually.

I think the moral philosophy is a lot more of a grey area. Those expounding in this area seem to likely be subject to the biases of their own beliefs and the particular fads of this time period just as much as everyone else.

Also, may I register my complaint to the Powers That Be about titles being edited yet again? I chose the title I did to highlight the aspect of the essay most of interest to HN readers; by editing the title to the original, you are partially responsible for the hostile reaction in the comments here where only one person actually replies to the content.

I think the Powers are beyond help. Battle not with SEO spammers, etc.

I disagree with the author's conclusions that fabs are immune from terrorist acts because a) fabs are too large, and b) terrorists are basically dysfunctional.

I can think of several relatively simple methods to take a large building out of commission.

And don't think for a second that terrorists don't know about all of them and more.

What do you disagree with from http://www.gwern.net/Slowing%20Moore%27s%20Law#state-actors-... ?

> I can think of several relatively simple methods to take a large building out of commission.

I'm sure you could, as could I - drive up a truck filled with fertilizer comes to mind. But that's not the question being asked, just attacking a random large building - for example, a truck bomb wouldn't work for a chip fab, though, which is relatively dispersed and well-guarded, and alternatives like 'send over one bomber and drop one daisy cutter on the complex' aren't available to terrorists.

> And don't think for a second that terrorists don't know about all of them and more.

Please provide 3 examples of competent terrorists using your methods.

Serious attempts to colonize space might be a tremendous complication here. There might arise a scenario where economics favoring the exportation of outdated fabrication equipment combined with the relative ease of constructing huge industrial facilities that are entirely in vacuum would result in a proliferation of low-power devices in the off-Earth economy.

Reply on news.ycombinator.com