• 0 Posts
  • 38 Comments
Joined 2 years ago
cake
Cake day: June 18th, 2023

help-circle

















  • I have almost the exact opposite view.

    I really like that much like a calculator or a very sophisticated search engine we can use large language models (LLMs) to enhance our cognitive capabilities. My memory works in such a way that I would be completely useless without a search engine, while I have almost unlimited capacity when it comes to puzzling together LLMs and search queries. I don’t feel that it takes anything away from humans, it’s just a tool like any other that allows us to expand our capabilities, just in a cognitive capacity!

    I would have never ever been able to figure out what that keyboard was called, but just by squeezing the answer out of a model that has obviously been trained on that data, we were able to figure out the answer.

    If it were to have been done by purely human memory, we would have had to have a person that actually knew this bit of super obscure knowledge and the same person by some amazing chance reading this post.

    I see no sadness, just a tool that is globally available to almost anyone (25 USD month for GPT4 or perhaps anyone being able to run their own open-source miqu-1–70b from mistral.ai) and has the capacity to elevate human cognition, simply by adopting a mindset of directing an LLM to get the results we need, much like typing a query into a search engine.