A global workspace in language models

Re:A global workspace in language models by josh-sematic on Monday July 06, 2026 @08:58PM (#48812441)

If I ask my kid (early elementary) what he learned last school year, I can watch him apparently genuinely trying to think of some examples and not being very successful. If I ask him something I know he learned he’ll gladly say “oh yeah!” and tell me about it.

[ Parent | on HN ]

Re:A global workspace in language models by michaelhoney on Monday July 06, 2026 @10:18PM (#48812957)

dude, you could ask me, a grown adult, what I did last week and get a similar lack of recall

[ Parent | on HN ]

Re:A global workspace in language models by nozzlegear on Tuesday July 07, 2026 @12:50AM (#48813774)

LLMs aren't humans, humans aren't LLMs. The fact that your kid can't recall what they learned without you reminding them is natural; the fact that the LLM can't recall something without a reprompt is a failure mode.

[ Parent | on HN ]

Re:A global workspace in language models by Ferret7446 on Tuesday July 07, 2026 @10:05PM (#48826627)

And yet (almost?) all of the quirks people keep bringing up about LLMs are also quirks that human cognition share.

It's almost as if artificial neural nets share emergent behavior with natural neural nets.

Pop quiz, what's the difference between natural synthesized vanillin and artificially synthesized vanillin?

[ Parent | on HN ]