Comments on: Thinking about ethical AGI, part 2

By: Shane Legg

Shane Legg — Wed, 15 Sep 2010 10:22:28 +0000

Not really. His writing sounds nice, and makes is seem like you’re understanding something difficult… but if you then try to go through what you’ve learnt all you find is vague intuitions.

Some topics are just hard, and you need to do hard work in order to understand them. Writing about the topic well can help a lot, but at some point the inherent complexity of some things simply must be faced if you’re really going to understand the subject.

By: Tim Tyler

Tim Tyler — Sun, 12 Sep 2010 21:02:18 +0000

*Not* a Hofstadter fan?!?

By: Snog

Snog — Sat, 07 Mar 2009 21:57:30 +0000

I sat next to him at a conference. The speaker asked who among the audience had signed up for cryro-presevation– requested a show of hands. E. Y. raised his with a scattering of others.

I admire him for raising his hand and I’d LOVE to see his head, slightly freezer burned, atop some surplus future-torso smiling and smug as ever, having anticipated the delightful impossibility of saying “I told you so,” to all the dolts like me who didn’t spring to have their heads preserved.

I do like his work.

A girlfriend with nipple rings would help uncramp his style.

By: Shane Legg

Shane Legg — Sun, 01 Jun 2008 21:42:07 +0000

I’ve only met him once in person, back in 2000 while I was working in New York. He have a talk on Friendly AI at our company. I only have a vague recollection of the event.

I’ve heard of flare… but don’t really know anything about it.

By: Dave Feltenberger

Dave Feltenberger — Sun, 01 Jun 2008 19:27:35 +0000

I incidentally went to Transvision 2007 in Chicago, where Yudkowsky spoke. One night I went out with him for dinner/drinks (along with several others), and aside from a razor sharp and hilarious wit he used with a borderline schizo Ayn Rand fan, he really rubbed me the wrong way. Literally he said without him a non-friendly AI will eventually take over the world. It sounds like the egoism may be slightly less intense now.

I hardly understand some of the things he writes as I’m not primarily interested in AI, so I can’t really comment on that. But have you seen the XML-based language, Flare, that he’s involved in creating? It’s intriguing – the mixture of static/dynamic typing, the Lisp-like data/program syntax, etc. It looks like the project never really took off the ground, though.