Can anyone point me to a good model policy on AI usage for open source projects?
I know the best policy is "fuck off, no AI" but that has no chance of getting past management. I need one that at least says you need to disclose LLM usage so that I can ban obvious bots or "people" who are just feeding things to ChatGPT.
There has been a notable increase in gen AI spam contributors lately and my previous policy of "just ignore LLMs" which worked for small/mid size projects is no longer effective for the mid size ones. I think we're approaching the point where because corporate/"higher education"/etc has been pushing it whole hog for years the younger generation doesn't even realize they're doing anything wrong.
The easiest way for a leftist to win an election is to run an unabashedly left wing campaign. This is because the vast majority of people want left wing policies.
The easiest way to lose an election is to listen to people who say otherwise.
The only way to understand most of it is to read the Linux src, and the only way to read the Linux src is to fight your way through 6000 layers of indirection.
I was just reading the Quectel AT commands manual and if it turns out the reason I couldn't get a response from the uart is because commands need to end with /r instead of /n I am going to be so mad/happy.
I put in a sim card and was able to send and receive a couple text messages with my wife. I was able to dial a phone call, but had no audio. It's possible there's some AT command to re-route the audio but I don't have time to look into it right now. This is definite progress.
Trying to find a Canadian online store to buy a usb uart cable from.
Pine store is Hong Kong based and it costs $2 but $30 shipping. Amazon is American and doesn't seem to do free shipping anymore to try and force you into getting prime. Costs $30 anyways. Canakit doesn't seem to have anything. Adafruit is American.
@sirjofri@mastodon.sdf.org's blog post about setting up dkim on 9front and @a@pdx.social's comment about adding a link to fediverse comments on it reminds me that I wanted to look into if parts of the dkim code could be extracted into an http signatures command that could be used for my plan9 activitypub implementation so that I could add activitypub support to my rc scripting based static site generator and wow I wish I wasn't so easy to nerd snipe.
Some guy in Montreal. Admin account for doomscroller.social.Mostly interested in #COVID, #aikido, #plan9 / #9front, #science and #neuroinformatics.#BringBackMasks