

Swede here.
Some American candy, mostly bad chocolate
Cryptography nerd
Fediverse accounts;
[email protected] (main)
[email protected]
[email protected]
Lemmy moderation account: @[email protected] - [email protected]
Bluesky: natanael.bsky.social
Swede here.
Some American candy, mostly bad chocolate
The judge explicitly did not allow piracy here. Only legally acquired media can be used for training.
This case didn’t cover the copyright status of outputs. The ruling so far is just about the process of training itself.
IMHO the generative ML companies should be required to build a process tracking the influence of distinct samples on the outputs, and inform users of potential licensing status
Division of liability / licensing responsibility should depend on who contributes what to the prompt / generation. The less it takes for the user to trigger the model to generate an output clearly derived from a protected work, the more liability lies on the model operator. If the user couldn’t have known, they shouldn’t be liable. If the user deliberately used jailbreaks, etc, the user is clearly liable.
But you get a weird edge case when users unknowingly copy prompts containing jailbreaks, though
Consider getting VoIP phone numbers from a jurisdiction that’s much less hostile, so you have another number available to use
Telegram also don’t have E2E encryption on groups
Massive singular radio telescopes are used to pick up individual signals from one direction, and can’t do imaging alone.
Sure you can pick up long wave radio with smaller antennas, but not without trade-offs. They often need long coils, and to make up the remaining difference you need to very precisely control electric resonance, and you lose efficiency (you pick up less energy from the radio waves). You definitely can’t do imaging with just one.
Just look at how big NFC and Qi coils are, they can’t practically be made smaller at those wavelengths, or else you lose too much energy!
Massive radio telescope arrays spanning the globe uses the massive distance to create a tiny amount of angular resolution, just enough that with months of processing you can image a black hole a few light years away with some thousands of pixels. Compare to how your phone can run deblur algorithms on a fraction of the power over far more pixels, because the angular resolution makes such a huge difference (blur radius is infinitely smaller)
Also, amplitude is signal strength. That’s only tall on a chart.
After checking for physical health issues, ergonomics, stress, posture, etc, take a look at better shoes. If you ever feel like your heels often hit the ground hard when you walk, and especially if you more often feel pain after having carried something heavy, the impact on your spine might be bigger than you think even if you don’t feel it at that moment. Better shoes (and posture, etc) reduces the impact significantly.
Maybe we could use this stuff for umbrellas too?
Not in the winter, lol.
I’d make them prove it, tell me something that happened or was said or seen in the vision that they couldn’t have known. Something which can be verified.
Chances are they can’t.
For public groups you probably want something like Matrix.org instead. Also open source, also supports E2E encryption
“please ponder your verification orb”
Closely related to the brain in a vat thought experiment
The standards are royalty free, so I’m not sure what that has to do with anything
Then you want them to advertise NIST PQ standards
… Which is also not necessary for single user password databases anyway
Every LLM is shit at dealing with version changes. They don’t understand it as a concept, despite all their training data.
*you’re
One shelter is very valuable, a second shelter has extremely questionable value to me.
I’m not sure you know what content addressing is.
There’s no definition of “inherently valuable” which doesn’t rely on arbitrary axioms. Especially because no amount of inherentness can guarantee a minimum price.
The Pixel line is comparable to the Samsung S line, you got a budget phone before