Thread

1. stratos123 · 1791107241 · focus · HN ↗

   LeCun also said back in 2022 that "if you train a machine, as powerful as it could be, your 'GPT-5000', on text", it will never be able to learn basic common-sense physics like that objects placed on tables will move along with them.

   1. mdp2021 · 1791117069 · focus · HN ↗

      > never be able to learn basic common-sense physics

      And has it at this stage, within in-depth take of said "learning", foundationally?

      I have not been able to properly check the studies for a long time now, but I remain unaware of achieved solutions on the problem of reliably referencing a world model out of a language model - that "counting the 'r's in 'raspberry'" be not guessing, not memory, but actually counting.

      1. ben_w · 1791133458 · focus · HN ↗

         To determine this, it would first need to be able to spell "raspberry" as letters rather than as tokens.

         Given you also don't want it to memorise [for all tokens, count([for all letters]), this would probably be more like "here's two images, count all things in the big image that look like the thing in the small image", which can then be r's in a photo of a raspberry jam jar in a supermarket, or dragons in a photo of a furry convention, or whatever.

         That said, they are competent enough at coding that I keep seeing them write code to do even simple tasks.

         On a related note: why did I see Claude editing a file by using cat to write a python script to do a grep search and replace?

         1. aesthesia · 1791140005 · focus · HN ↗

            > Given you also don't want it to memorise [for all tokens, count([for all letters])

            Why not? You've memorized how words are spelled, and how sounds correspond with letters, and how concepts correspond with words. To the extent that there are shortcuts that enable compression you use these, and the model will do something similar.

            1. ben_w · 1791142744 · focus · HN ↗

               Combinatorial explosion, and facts merely memorised is a huge waste of parameters that are better dedicated to effective reasoning.

               Being able to spell all the words then count letters is simpler, and more generalisable to other tasks, than memorising answers to all possible word questions.

               1. aesthesia · 1791152763 · focus · HN ↗

                  Ah, I misunderstood what you meant. I was just trying to highlight that in order to answer these types of questions the model needs to memorize the spelling of each token. But you're right that that's all they need to memorize, and algorithms like counting are pretty simple for transformers to implement.
            2. mdp2021 · 1791147070 · focus · HN ↗

               > Why not?

               Because to "123x456" we want a reply that goes "this times that plus that...", not "Was that not nnnnnn?". If it does not perform its duty it is a liability.

         2. mdp2021 · 1791146845 · focus · HN ↗

            > it would first need to be able to spell "raspberry" as letters rather than as tokens

            Of any object in question they should be able to create a representation that allows correct assessment.

            > Given you also don't want it to memorise

            That is obviously necessary: what we want from the consultant is to check, not to remember. Answers must be correct and that implies having performed all due diligence - and being capable of doing it, before that. So, objects must be instanced internally in a way that allows effective handling. Counting letters is a good example of the ability (that must remain general).

Unofficial Hacker News client; not affiliated with Y Combinator.