You're looking at Less Wrong's discussion board. This includes all posts, including those that haven't been promoted to the front page yet. For more information, see About Less Wrong.

Luke_A_Somers comments on Open thread, January 25- February 1 - Less Wrong Discussion

8 Post author: NancyLebovitz 25 January 2014 02:52PM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (316)

You are viewing a single comment's thread. Show more comments above.

Comment author: Luke_A_Somers 26 January 2014 03:15:03PM 2 points [-]

Heretical? Well, considering that 'heretic' means 'someone who thinks on their own', I'm not sure how we're supposed to interpret that negatively.

I assume however that you meant 'disagreeing with common positions displayed on LW' - which of those common positions did you differ on, and why, and just how homogeneous do you think LW is on those?

Comment author: Thomas 26 January 2014 04:20:15PM *  0 points [-]

I can speak mostly for myself. Still, we the locals go back decade and more, discussing some topics.

It is kind of clear to me, that there is a race toward superintelligence. As it was always the race toward some future technology, be it flying, be it atomic bomb, be it Moon race ... you name it.

Except, that this is the final, most important race ever. What can you expect then from the competitors? You can expect them to claim, that the Singularity/Transcendence is still far, far away. You can expect, that the competition will try to persuade you to abandon your own project, if you have any. For example, by saying that an uncontrollable monster is lurking in the dark, named UFAI. They will say just about anything, to persuade you to quit.

This works both ways, between almost any two competitors, to be clear.

My view is the following. If you are clever and dare enough, you can write a 10000 lines or there about long computer program, and there will be the Singularity the very next month.

I am not sure, if there is a human (group) currently able to accomplish this. Very well might be. It's likely NOT THAT difficult.

We discussed the Marylin vos Savant's toying with Paul Erdos. A smartass against a top scientist is occasionally like a cat and mouse game, where the mouse mistakenly thinks he's a cat. There are many other examples, like Ballard against all the historians and archeologists. Or Moldbug against Dawkins.

Of course, that does not automatically mean another smartass is preying upon the MIRI and AI academia combined, in the real AI case. But it's not impossible. May be several different big cats in the wild who keep a low profile for a time being. Might be lion with his pride, inhabiting the academia also.

The most interesting outcome would be no Singularity for a few decades.

Comment author: Lumifer 26 January 2014 05:03:13PM 3 points [-]

If you are clever and dare enough, you can write a 10000 lines or there about long computer program, and there will be the Singularity the very next month.

That seems an... unusual view. Have you actually tried writing code that exhibits something related to intelligence?

10K lines is not a big program.

Comment author: Luke_A_Somers 27 January 2014 03:57:18PM *  0 points [-]

It depends on your language and coding style, doesn't it? I've seen C style guides that require you to stretch out onto 15 lines what I'd hope to take 4, and in a good functional language shouldn't take more than 2.

Comment author: Lumifer 27 January 2014 04:18:16PM 1 point [-]

Yes, and the number of lines is a ridiculously bad metric of the code's complexity anyway.

Was a funny moment when someone I know was doing a Java assignment, I got curious, and it turned out that a full page of Java code is three lines in Perl :-)

Comment author: Luke_A_Somers 27 January 2014 04:22:04PM *  0 points [-]

That really depends on coding style, again. I find that common Java coding styles are hideously decompressed, and become far more readable if you do a few things per line instead of maybe half a thing. Even they aren't as bad as the worst C coding styles I've seen, though, where it takes like 7 lines to declare a function.

As for Perl vs Java... was it solved in Perl by a Regex? That's one case where if you don't know what you're doing, Java can end up really bloated but it usually doesn't need to be all that bad.

Comment author: Lumifer 27 January 2014 04:40:13PM 0 points [-]

As for Perl vs Java... was it solved in Perl by a Regex?

I don't remember the details by now, but I think that yes, there was a regexp and a map, and a few of Perl's shortcuts turned out to be useful...

Comment author: Thomas 26 January 2014 05:22:50PM 0 points [-]

I have certain abilities. This is the product of the product of mine from 10 years ago.

Smartass I am. Probably not smart enough to really make a difference, though.

Comment author: Lumifer 26 January 2014 06:04:20PM 3 points [-]

Smartass is good. Saying things which are clearly not true without a hidden smartassy implication behind them -- not so much :-)