• 0 Posts
  • 4 Comments
Joined 3 years ago
cake
Cake day: May 11th, 2024

help-circle


  • I completely get you, man.

    During a “AI traning” my job enlisted me, the instrutor, someone with apparently high technical knowledge of software systems and development, design patterns, with years of experience and supposedly familiar with using CLIs and Linux shell commands wanted to create a symlink to a file and instead of typing up the command, or if he did not remember it, just quickly looking up on a usual search or even a LLM powered search which command to use to archieve it, he wrote up on his Claude Code terminal “create a symbolic link from X to Y”…

    The Claude Code agent of course performed the action correctly, however the amount of processing power wasted to execute some basic action of his job, which is executing a command correctly, was all left for the “AI” agent to do because it was “easier”.


  • Yes, but most of the paid “LLMs” are not actually just LLMs, they have specific aritimetic modules that end up being activated for most math questions that actual LLMs suck at, however are pretty easy for arithmetic modules, like simple calculations, unit conversions, etc…

    So yeah, any edge “LLM” tool currently will always get the original question asked here CORRECT even on the free tiers, you can test it yourself on Google Search itself in “AI mode”, on ChatGPT, or DeepSeek…

    So you are technically correctly in saying that LLMs suck and cannot be trusted for math in almost any case. However you are wrong on implying that ChatGPT, or other commercial models are just LLMs, they are not.

    P.S.: I’m not condoning the use of AI to resolve math related questiona, nor saying that their replies should be trusted, just stating the fact that actual “AI” products use more than just a large language model to generate their responses, and for math related questions most “AI” today will get arithmetic question answered correctly NOT DEPENDING on having training for the exact questiona being asked since they can identify a math arithmetic question being asked and instead of trying to fully generate the answer with the model itself, which would lack the training data to be able to process such wide range of possible question, they generate placeholders internally and let a arithmetic like module do the calculations, which take negligible processing power to be cacoulated by CPUs.