tinwhiskers

One of the things that weighs me down is posts making me dwell on the things that weigh me down.

“Despite being so common in English as to be known as the “Chinese curse”, the saying is apocryphal, and no actual Chinese source has ever been produced.” - Wikipedia

I’m located in a van in New Zealand so I only use mobile data. I pay NZ$40 (US$25) per month for “unlimited” data, which is all I can eat but capped at 1Mbps. I can stream 720p barely, but I mostly torrent. I typically use about 60-80GB a month.

OruxMaps (android) supports several navigation methods, kml overlays, offline maps, various online and custom maps, good tracking, routes, gps, etc, etc. Waaay better than Google maps - although it can also happily use Google maps.

I think you’re right there. My bad.

They say the second layer retains 93% of the performance of the first using reflected light, making it 20% efficient, so, yes they are added in that case.

TLDR; the front side is 23% efficient, and the rear side 20% efficient.

They don’t actually give an overall efficiency but it implies a total of 43%. They compare this to typical panels also at 23% efficient, so it’s really remarkable if true. Other emerging solar tech is up to about 32% but if that could also benefit from multiple layers then total efficiency could become insane.

Seems a little too good to be true, really, but great if so.

Edit: Yeah, I don’t think these efficiencies can be added like that. I guess the overall efficiency will depend on how reflective the ground under the panels is, and they will extract 20% of that. Maybe that’s why they don’t give an overall rating.

I think it’s intellectually lazy to stick with the stochastic parrot line of thinking now. There’s a number of emergent properties that are appearing as LLMs scale that give them abilities beyond that paradigm. Check out the “Sparks of AGI” paper from Microsoft research - or more realistically one of the youtube summaries of it since its quite a big read… Here’s one from the horse’s mouth: https://m.youtube.com/watch?v=qbIk7-JPB2c

Anyone who isn’t at least mildly interested that you know Morse code isn’t someone you want to know :-)

Good filter technique.

No. I only use Android as my PC via AR glasses. Is there even any antivirus software for Android? Probably, but I don’t care I guess. Never had a problem.

FYI, you are able to edit titles on the fediverse in case you want to the missing word.

Well of course, putting it on the open internet is very intentionally making it available for everyone to see. If you don’t want everyone to see it, don’t put it on the open internet. The issue is what people do with it, not whether they can access it. Copyright forbids distributing copyrighted data. The entire point of that it is so that you can make it available to be seen but protected from people copying it. However, there is no distribution or storage of copyrighted material with an LLM - there is no copy. I think OpenAI will be OK, but these things are never certain when the big lawyers are let loose.

Distributing the training dataset, though, that could well be a problem.

I’ve been using signal for a few years. I’m now in a situation that I use two different phones but signal only allows you to be connected to a single device, and you only see the messages in your history for the particular device you received it on. You can’t migrate messages, the UI is extremely basic and isn’t a responsive design so you have no control over window size on larger screens. There’s just so many annoying little things about it, but overall it’s pretty solid. I’m looking for a better solution now though.

It really needs some more effort put in on UI design, data migration and linked devices. Development is very slow. I hoped it would improve but nothing has changed for years.

The fediverse wars, begun they have.

My take on it was that it was just flag you can set to openly identify the account as a bot, but nothing to prevent either a human or bot from lying. There’s bot disclosure laws in progress in the EU and California in some form, so it may be future proofing, too.

Yeah, I just signed up to lemmy after using kbin for a couple of days. The problem described is really annoying.

Don’t sign up to lemmy.ml or you’ll need to sign up again on a new instance when they get defederated. Tankies. I guess a blind eye is being turned right now because numbers are important for survival atm, but I wouldn’t consider it a future-proof instance.

Yes this is exactly what happened with the freenode -> libera split. Although we entirely lost some good people and the community was smaller, the reduction in toxicity made it all worthwhile. The type of people who defend reddit on this are not the type of people we want anyway.

This is the first real chance we’ve had to break the chains on corporate social media. We should all strive to make this work for the betterment of everyone in the longer term.

Much the same happened with the Slashdot -> soylent split but the community was already so small by then it had no chance of thriving. We don’t need a huge chunk of reddit to survive and thrive though.

Yes, it’s working again.

You could try kbin. It has an easier path to entry and being in the fediverse still able to access the same content. I do not intend to start a fediverse war here though.