3 Matching Annotations
  1. Nov 2025
    1. we can’t recapture the same processes we used to learn to speak for the very first time

      for - unlearning language - key insight - language - cannot recapture same process we used as child - cannot recapture the same processes we used to learn to speak language for the very first time - basically, we lose access to that original vocal learning circuit as an adult - question - language learning - what is this vocal learning circuit of an infant? - why do we lost access to the vocal learning circuit we had as a child? - observation - clue - language - accidental world recall and substitution - a clue to how we remember words - I wrote the above sentence "why do we lost access to the vocal learning circuit we had as a child?" when I meant to write: - "why do we LOSE access to the vocal learning circuit we had as a child?' - This very observation also has the same mistake: - "observation - clue - language - accidental world" instead of: - observation - clue - language - accidental WORD"! - I've noticed this accidental word substitution when we are in the midst of automatically composing sentences quite often and have also wondered about it often. - I think it offers an important clue about how we remember words, and that is critical for recall for using language itself. - We must store words in clusters that are indicated by the accidental recall

  2. Jun 2022
    1. "The idea is that over long periods of time, traces of memory for visual objects are being built up slowly in the neocortex," Clerkin said. "When a word is spoken at a specific moment and the memory trace is also reactivated close in time to the name, this mechanism allows the infants to make a connection rapidly." The researchers said their work also has significant implications for machine learning researchers who are designing and building artificial intelligence to recognize object categories. That work, which focuses on how names teach categories, requires massive amounts of training for machine learning systems to even approach human object recognition. The implication of the infant pathway in this study suggests a new approach to machine learning, in which training is structured more like the natural environment, and object categories are learned first without labels, after which they are linked to labels.

      visual objects are encoded into memory over a long period of time until it becomes familiar. When a word is spoken when the memory trace associated with the visual object is reactivated, connection between word and visual object is made rapidly.