On Tuesday, May 26, 2020, at 11:30 AM, Matt Mahoney wrote: > But this is not a path to AGI. The child program doesn't know any more than > the parent. Matt seems to have made a obvious "typo". More data/ compute makes you smarter, but the algorithm that can extract more/better data and do so faster is smarter because it can get smarter faster. The "previous current AGI" can learn how to make such new AGI.
BTW those linked papers are wordsalad nonsesense, are not clear, and have very little useful information in them it seems (lots of filler words). Don't confusify AGI, explain AGI like its a toy. My "paper" is almost the same length and has 50x more information than the one above. If we don't unify many things and single them out too much, we won't be efficient. ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T5364f6a30c9294ea-M7574adb4b763777f92bbc7ce Delivery options: https://agi.topicbox.com/groups/agi/subscription
