3 ms·
My understanding of AM and Eurisko (having looked into them a decade or so ago) was that their source code hadn't been published, and that there was a dispute a
by cabalamat 3y ago
My understanding of AM and Eurisko (having looked into them a decade or so ago) was that their source code hadn't been published, and that there was a dispute as to what their capabilities actually were and how much was exaggeration by Lenat.
I don't know if that's still the case. I do think that it would be worth creating systems that mix the ANN and GOFAI approaches to AI.
- earleybird 3y agoHad to go digging . . . Kenneth Haase did a Phd (1990) where he builds an exploratory system called Cyrano and does a substantial review/analysis of AM (& Eurisko) https://dspace.mit.edu/handle/1721.1/14257?show=full https://dspace.mit.edu/handle/1721.1/14257?show=full It's quite interesting.
- cabalamat 3y agoIt may well be, but when I try to download the pdf it fails.