Diary 2023-04-10
@Developer65537: I've been using ChatGPT to compress the number of tokens. I've tried using it, and it's certainly very powerful and worth pursuing when creating a service like Zenryoku Katsudo Boyfriend-kun, which is about how much you can remember about the user experience (how much information you can pack into a small number of tokens)... ↓Example of token count reduced by a factor of 6
https://pbs.twimg.com/media/FtPh_z1akAEN5Mg.jpghttps://pbs.twimg.com/media/FtPh_0maEAAWFJd.jpghttps://pbs.twimg.com/media/FtPh_z3aAAAB2-P.jpghttps://pbs.twimg.com/media/FtPh_zvaQAI-F9l.jpg
https://gyazo.com/83dfc96f333c7285ac219229f0d3e5a3
Hmmm, I've tried to follow up, but it doesn't restore that well.
@todesking: horror (it's not restored at all and people don't even know it) This interpretation is more likely to be correct.
The input and output ports of information need not be the same.
The concept of a "social networking account" only temporarily produces "a bundle of inputs and outputs."
Biologically, the ear and mouth are different, and even before the Internet, reading and writing were two different things.
If so, is it natural for social networking accounts to have separate ears for information input and separate mouths for output?
↔ In some cases, output increases input.
---
This page is auto-translated from /nishio/日記2023-04-10 using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I'm very happy to spread my thought to non-Japanese readers.