3 ms·
> Puzzles a child can do Certainly, I couldn't solve Hanoi's towers with 8 disks purely in my mind without being able to write down the state of every step or
by neoden 1y ago
> Puzzles a child can do
Certainly, I couldn't solve Hanoi's towers with 8 disks purely in my mind without being able to write down the state of every step or having a physical state in front of me. Are we comparing apples to apples?
- ben-schaaf 1y agoWriting things down and reading them back is quite literally the only thing LLMs do.
- neoden 1y agoGenerating text into the current context is not the same as writing down. It's the same as having a thought and putting it into short-term memory. An analogy for writing down would be sending something to an MCP server that provides context-independent memory functionality.
- RugnirViking 1y agoWhy without writing down each step? Would you be able to solve it writing each step required in sequence? Thinking between each one? Pretty sure i could, isn't that closer to an LLM?
- neoden 1y agoI mean I need to offload state of the puzzle being solved from my brain to an external memory device — paper, in this case. Keeping that state in my mind would be much harder. It's like some people can play chess within their minds without a board, but it's obviously not something that everyone can do
- RugnirViking 1y agoBut the LLM can think between writing each token, and indeed can factor in it's own previously written tokens into its answer - thats essentially using a piece of paper and writing stuff down and referring back to it. Thats the whole idea behind thinking models, and they are demonstrably better at many tasks than others.
- neoden 1y ago> But the LLM can think between writing each token Writing a token is the thinking itself. Thinking models just write some tokens behind the scene, that's the whole difference.