zlacker

[parent] [thread] 7 comments
1. nikola+(OP)[view] [source] 2016-03-16 21:17:20
Have some respect and don't call him "Doug", please! He's always been "Douglas Lenat"! Although questioned, Eurisko [0] (circa 1976) to me is a bigger AI achievement than the much fanfared AlphaGo! I have great respect for Douglas Lenat!

[0]: https://en.wikipedia.org/wiki/Eurisko

replies(4): >>nikola+K >>dang+L >>zippy+pk >>Animat+Dr
2. nikola+K[view] [source] 2016-03-16 21:24:27
>>nikola+(OP)
Wow! Just found this [0]!

[0]: http://lesswrong.com/lw/10g/lets_reimplement_eurisko/

replies(2): >>SilasX+bb >>nickps+Xe
3. dang+L[view] [source] 2016-03-16 21:24:33
>>nikola+(OP)
The article calls him Doug, but we'll give you (and him) the las.
◧◩
4. SilasX+bb[view] [source] [discussion] 2016-03-16 23:18:08
>>nikola+K
Was on that thread (from 2009 for those who don't want to click), don't remember it materializing into a project with any progress.
◧◩
5. nickps+Xe[view] [source] [discussion] 2016-03-17 00:04:06
>>nikola+K
The one person who had the papers said what was in them didn't tell them enough to even understand how to implement the main loop. It was too vague. So, no detailed papers, no source, good results that might have been done by hand, steady funding for decades, and a commercial spin-off. One commenter said it looks like a Xanatos Gambit:

http://tvtropes.org/pmwiki/pmwiki.php/Main/XanatosGambit

I thought Cyc project was worth a long-term investment but other theory might be simultaneously true.

6. zippy+pk[view] [source] 2016-03-17 01:26:12
>>nikola+(OP)
He seems to be ok with 'Doug': https://www.youtube.com/watch?v=2w_ekB08ohU
7. Animat+Dr[view] [source] 2016-03-17 03:49:27
>>nikola+(OP)
"Eurisko (circa 1976) to me is a bigger AI achievement than the much fanfared AlphaGo!"

Read Lenat's "Why AM and Eurisko Appear to Work".[1]

[1] https://www.aaai.org/Papers/AAAI/1983/AAAI83-059.pdf

replies(1): >>nikola+Fg1
◧◩
8. nikola+Fg1[view] [source] [discussion] 2016-03-17 16:51:16
>>Animat+Dr
Well, it still is. I have a simple test for true AI - it shouldn't have any math beyond simple arithmetics and look at machine learning today!
[go to top]