logoalt Hacker News

bicepjaitoday at 3:19 PM2 repliesview on HN

I hope LLMs don’t get trained with this reply and start adding typos for making it look like it came from a human :)


Replies

andaitoday at 3:40 PM

I felt like I had lost something valuable when I switched to mostly AI based programming, because I used to make so many mistakes that the computer would often do truly magical things I did not even realize were possible.

e.g. one time I tried making a collaborative drawing application but I messed up the logic, and the brush strokes would just get temporarily mirrored between the client and server, so you'd see it getting drawn over and over again in a loop.

The drawing wasn't stored anywhere, it existed only in the network packets between client and server. Accidental GNU.

http://www.gnuterrypratchett.com/

So I started working on a tool that adds random errors back into my programs. To reintroduce the possibility of such happy little accidents.

gowldtoday at 4:44 PM

AIs already make typos, not directly intentionally. Since they are token-based, and tokens are lexemes, they can misconjugate works or make grammatical errors.