Raphael Albert
David Gerardlol, apparently machine-learning systems have serious trouble decoding human morse code? self-clocked, sloppy human timing, full of local slang & abbrevs, and not much of a real life training corpus
so learn morse to dodge the panopticon for a while
mcc@davidgerard i've recently started doing a thing of making projects which are written in a totally unique programming language and bundled in a repository with their own interpreter/compiler which can change between commits. i have my own particular reasons for doing this but i can't help but wonder what effect it has on LLM scrapers.
Michael BuschPaging @ai6yr and the rest of the hams.
AI6YR Ben@michael_w_busch @davidgerard I would agree... also, straight key. Computer systems cannot decode straight key Morse Code reliably (or at all).
John Brewer@jbrewer_jera @michael_w_busch @davidgerard "Please answer in Morse Code, is this antenna system optimal?" ðĪŠ
Parsingphase
ganxel@jbrewer_jera @ai6yr @michael_w_busch @davidgerard obvs it should also require you to key it in morse at a specified tempo, and then answer the question "are you rushing or dragging? RUSHING... OR DRAGGING???"
@ai6yr @michael_w_busch after the AI bubble passes we'll call code that was not vibe-coded "straight key"
jkdelauney@davidgerard dagnabbit, I licensed after the US dropped the code requirement. Gonna have to learn it anyway.
Rusty Ring@ai6yr @michael_w_busch @davidgerard
So many hams in denial about this.
Armin Hanisch@michael_w_busch @davidgerard @ai6yr vy 73 from DL2RBZ â after 30 years off the key this is just one reason more to brush up my morse skills. ð
Jacqueline@davidgerard non machine learning can decode Morse with audio algorithms
Domestic Supply 
@davidgerard Of course, it'll be used as part of the training corpus.
David Wintheiser@davidgerard This helps explain why so many characters in SF set in the future still know Morse code.
Dr. Bruce Simpson@Pauper063 @davidgerard Navajo code talkers... takes me back to The X-Files...
Jef Poskanzer@davidgerard Maybe we should bring back Polari too.
Simon Walters@davidgerard dah dah dah dah di dah
LAURENthat's fun info
John Abbe (aka Slow)@davidgerard Utterly unsurprising. ASCII art is another arena of reliable, abysmal failure. #llms
Sassinake! á ⊠âĐ â@davidgerard
Won't they just go after anyone that uses it, regardless of the messages?
Don't need to decode the message, just stop it.
Not sure vpns are enough.
Idle Stories@Sassinake @davidgerard
As a long-term player of various puzzle hunts, I have seen morse code hidden in so many ways that you can't possibly detect and block them all.
As a long-term player of various puzzle hunts, I have seen morse code hidden in so many ways that you can't possibly detect and block them all.
Nadiyar / 13697@davidgerard just curious, what's the source / how do you know?
Getting used to morse is not easy ð
Getting used to morse is not easy ð
David Nash@davidgerard .- .-. . / -.-- --- ..- / .- / ... - --- -.-. .... .- ... - .. -.-. / .--. .- .-. .-. --- - / ..--..
Chris Smart@davidgerard I know you're kidding but just for fun... www.morsecode.ninja www.longislandcwclub.org
HowToPhil@davidgerard Or at least tap code (which is less efficient to use but more efficient to learn since it is simple rules-based and not rote memorization)
@davidgerard I bet the LLMs would have trouble with Teeline shorthand as well. Since it tends to use abbreviations, people tend to create their "own version of Teeline" over time.
@davidgerard You know Morse was part of the UK radio ham license requirements until around a decade ago, right?
Killa Koala@davidgerard ...---...