3 ms·
Writing is fundamentally the transfer of information from your brain to my brain. If you have 1000 bits of information you want to transfer, you can't give 300
by hatthew 7mo ago
Writing is fundamentally the transfer of information from your brain to my brain. If you have 1000 bits of information you want to transfer, you can't give 300 bits of information to an LLM and have it fill in the remaining 700, because it doesn't know what those 700 bits are. If it's able to guess those 700 bits correctly, then they aren't true information, and you really only have 300 bits you want to transfer. You might as well transfer those bits to me directly, rather than having the LLM add on an extra superfluous 700 bits that I then have to filter out.
- TZubiri 7mo agoRemember when LLMs were entering the mainstream, and everyone shared tips on how to superprompt, and one of the hot tips was to tell the AI to write a prompt itself? Like: BAD: "Write the specs for a system that does X Y and Z" GOOD: "Write me a prompt to write the specs for a system that does X Y and Z." As if the LLM magically knew all about itself and the best tips and tricks to prompt itself even if it had just came out and there was no scraped information on how to use them yet.
- charcircuit 7mo agoIt's not possible to directly transfer 300 bits of information as that would require a form of telepathy. If it takes 1000 bits of English to distribute something that represents your idea properly it is useful to have a tool to create those 1000s bits instead of doing it manually character by character.
- hatthew 7mo agoSorry I wasn't very clear. TFA is talking about using LLMs to write things from scratch, not just to clean up grammar for example. In that context, I was talking about bits of semantic information, not bits of English text information. You might have 300 bits of semantic information in your mind, and then you have to expand that to, say, 600 bits of English text to give to the LLM. If you're using the LLM purely to turn bullet points into prose, it'll add more bits of English, but not more bits of (useful) semantic information. I prompted Claude with "(information theory) difference between semantic information and english text information in the context of using LLMs for writing": https://claude.ai/share/5925245a-0893-46ba-bca9-30627d4facbc https://claude.ai/share/5925245a-0893-46ba-bca9-30627d4facbc If you're familiar with LLMs and information theory, the LLM isn't giving you any semantic information that you don't already know. If you aren't familiar with LLMs and information theory, you can learn about them from google and/or your own LLM, using that prompt for keywords. In either case, the LLM's response isn't very helpful, because it's not my ideas that you are reading, it's random information pulled from the internet (directly or indirectly), and it's not actually the semantic information I wanted to convey. This comment is more useful than the LLM's, because every word is chosen to convey the ideas in my mind as clearly as possible in the context of this article and conversation. It's also half as many words to read.
- charcircuit 7mo agoI do think that a list of bullet points versus an article give a different impression to the reader. The same information packaged in different ways can give a different impression and I think that impression is part of the message people want to give. Reading a prompt of bullet points + desired impression will give the reader a different impression than what the LLM would output.
- hatthew 7mo agoThat is very true! Subjectively, I would much prefer to read either bullet points or the impression you want to convey. I don't care what impression the LLM wants to convey.