I assume you don't mean magic and do understand what GPT "is". As for the non storing/updating weights, well you need to store memories in a tree at least, if not a more merging tree like GPT networks (word2vec etc). Because it allows much faster access. Second, learning/updating weights is the sameish thing, you build on the tree using memories, which would otherwise take long to gather every time you want to answer a problem. ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/Tf6dddebe1e89183a-Mee826dc61d62314574c08772 Delivery options: https://agi.topicbox.com/groups/agi/subscription
- Re: [agi] Re: Advanced robots Nanograte Knowledge Technologies
- Re: [agi] Re: Advanced robots Nanograte Knowledge Technologies
- Re: [agi] Re: Advanced robots immortal . discoveries
- Re: [agi] Re: Advanced robots Matt Mahoney
- Re: [agi] Re: Advanced robots Nanograte Knowledge Technologies
- Re: [agi] Re: Advanced robots Matt Mahoney
- Re: [agi] Re: Advanced robots Ben Goertzel
- Re: [agi] Re: Advanced robots immortal . discoveries
- Re: [agi] Re: Advanced robots Ben Goertzel
- Re: [agi] Re: Advanced robots Rob Freeman
- Re: [agi] Re: Advanced robots immortal . discoveries
- Re: [agi] Re: Advanced robots Quan Tesla
- Re: [agi] Re: Advanced robots stefan.reich.maker.of.eye via AGI
- Re: [agi] Re: Advanced robots immortal . discoveries
- Re: [agi] Re: Advanced robots Nanograte Knowledge Technologies
- Re: [agi] Re: Advanced robots magnuswootton81
