You are reading a single comment by @NotThamesWater and its replies. Click here to read the full conversation.
  • I think I have said this before, but I find it fascinating as I think it is almost unfiltered stream of consciousness.

    If you can get some context from elsewhere, it's fairly direct steps from one thought to another, but there are rarely any internal checks that most people consider essential (but carry out automatically):

    • whether a branching idea should be considered at all

    • prioritisation of ideas (or even ability to store more than one or two simple ideas in the 'foreground')

    • considering whether an idea should be spoken at all, further processed or dropped

    It's kind of like his brain is a single core CPU with very limited storage that can only be flushed to voice.

    This makes it similar to output of a basic natural language generation algorithm, e.g. using a Markov chain.

  • This makes it similar to output of a basic natural language generation algorithm, e.g. using a Markov chain.

    He speaks by using autocomplete.

About