Skip Navigation

Posts
0
Comments
330
Joined
4 mo. ago

  • i feel like if it's not something thst everyone got this would cause new kind of racism

  • It's been a thing for years, way before api shitshow

  • helical antennas work fine too and look goofy as hell

  • i've used the same (800ml can) and this one works well. cookie tin is 15cm dia 8cm tall and it works, but size can vary a bit. you can copy or scale slightly designs of 13cm band antennas

  • wait i thought for some reason that pringles can sized waveguide would have cutoff frequency above 2.4ghz. nevermind, there's something better because bigger aperture can get you more directivity like this https://lea.hamradio.si/~s53mv/wumca/cup.html i made two out of cookie tins and it works over 500m at least

  • pringles can is too small for 2.4ghz cantenna, it's near cutoff frequency but just barely, you need 10cm-ish diameter can or shorter 16cm-ish can

  • if it's opt-out, it also keeps use of the shitty ai dubbing high thus making it an artficial use case. it's like with gemini counting every google search as single use of it

  • aliexpress did that since forever but you can just set display language once and you're done. these ai-dubs are probably worst so far but can be turned off by uploader (it's opt-out) (for now)

  • why doesn't anthropic, the bigger startup, simply eat anysphere?

  • if someone is so bad at a subject that chatgpt offers actual help, then maybe that person shouldn't write an article on that subject in the first place. the only language chatgpt speaks is bland nonconfrontational corporate sludge, i'm not sure how it helps

  • in one of these preprints there were traces of prompt used for writing paper itself too

  • maybe it's to get through llm pre-screening and allow the paper to be seen by human eyeballs

  • maybe there's just enough text written in that psychopatic techbro style with similar disregard for normal ethics that llms latched onto that. this is like what i guess happened with that "explain step by step" trick - instead of grafting from pairs of answers and questions like on quora, lying box grafts from sets of question -> steps -> answer like on chegg or stack or somewhere else where you can expect answers will be more correct

    it'd be more of case of getting awful output from awful input

  • nah, what happened is that they were non-psychotic before contact with chatbot and weren't even usually considered at risk. chatbot trained on entire internet will also ingest all schizo content, the timecubes and dr bronner shampoo labels of the world. learned to respond in the same style, when a human starts talking conspirational nonsense it'll throw more in while being useless sycophant all the way. some people trust these lying idiot boxes; net result is somebody caught in seamless infobubble containing only one person and increasing amounts of spiritualist, conspirational or whatever the person prefers content. this sounds awfully like qanon made for audience of one, and by now it's known that the original was able to maul seemingly normal people pretty badly, except this time they can get there almost by an accident, getting hooked into qanon accidentally would be much harder.

  • No. Barrels of API (active pharmaceutical ingredient) are mostly hauled from India or China, then formulated into pills or whatever, this goes especially for generic medicines. There is some american manufacture of APIs, but these tend to be on more expensive side (biologicals or small molecule drugs under patent). Inputs for these APIs also tend to be made in India or China

  • musk voice if machine god didn't want me to fuck with the racism dial, he wouldn't make it