then they'll find some sort of technical excuse and pull the plug on ActivityPub support
How do they do this without running a foul of regulators? People are already mad at meta and want to break them up for having instagram and Facebook, if they add the last big social media platform every politician right and left will be lining up to take them down. There's a reason they never bought twitter despite being able to 10x over. Combine that with new EU interoperability laws and there's no way meta could get away with that.
Washington stayed with his troops but kept behind the lines. Western Leaders mostly stopped leading from the front after Gustavas Adolphus got killed doing so.
The platform never really took off. It was a niche messaging platform before Facebook and Google and went back to being one after they left. I have yet to see any evidence that Google or Facebook helped or hurt xmpp, just speculation and anger that it didn't take off.
What is a government and what is a terrorist group, is Hezbollah a government or a terrorist group?, was the Jim crow bourbon democratic party a government or a terrorist group?, is the Iranian state a government or a terrorist group?
Governments and terrorist groups aren't mutually exclusive. Many governments, including the U.S., fund and support terrorist groups and terrorist acts, and many terrorist groups provide social and governmental services. Hamas does both, although there terror wing is where most of the resources and energy go, they still are, or at least were, the government of gaza and provided some governmental and social services. Them being a government doesn't make there atrocities any better or worse, same with the Israeli state.
To all the people wondering about metas intentions in this it's not the big bad corporation taking down the upstart competition. All the people saying it's EEE can't show any sign metas doing this or even wants to because the strategy doesn't work, any time a company does it it either doesn't take off or they get brought up on anti-trust laws. Show me a standard that was destroyed by EEE and I'll show you a standard that never took off in the first place. All the usual examples given, email, java, html, remain open standards to this day.
The truth is the fediverse isn't competition to meta, it's a fraction of the size and is populated by users who would never use meta services in the first place. They can pretend it's a competitor though. If twitter does actually collapse and people switch to threads meta will face anti-trust suits for owning the three largest social media platforms. If they add activity pub support though they can point to the fediverse and say it's competition, even if it's only 1 % of the platform. They also have to deal with EU interoperability laws that might start getting enforced.
TL;DR this is about compliance for meta, not conquest.
Calling llm "big auto-complete" is like calling people "big bacteria" . It's true that they act on the same goal, guess the next word for llm and auto-complete; survive and reproduce for people and bacteria, but they are vastly different in scale and complexity.
Also what would AI be to you if not an llm? Cause I'd say anything that has an SAT score higher than most Americans has some form of intelligence.
So is it alright if Russian civilians are bombed because of their broad support for Putin?, or u.s. and Israeli citizens get bombed for their overwhelming support for the IDF and their atrocities?
Killing civilians is bad no matter their politics.
You know who you don't see constantly complaining about retail theft, grocery stores. Probably because they have a business model resistant to the real cause of all these losses, online shopping and the decline of retail.
It's easier for the execs though to blame it on retail theft and tell their shareholders that they're gonna lobby Congress to lock up shoplifters and solve the problem, rather than tell them the business is slowly dying and there's nothing they can do about it.
Western nations give into terrorist demands going back to the French revolution. Some of those demands were for the freedom of speech that is being trampled on here or other rights and protections we hold dear. For recent examples look at the troubles or even that guy who shot shinzo abe and got the moonies out of Japan.
The focus shouldn't just be on the means for political change, though the means can be criticized, but the political change itself. Banning book burnings in this case is an afront to free speech and should not be implemented.
Agree mostly but I don't think unions are the solution to over-consumption. They're good for income inequality worker safety and a bunch of other things, but they're just as incentivized to increase consumption as corporations. Consumption is good for workers since it guarantees job security and possible increases to wages. To counteract these interests you need something representing the interests of everyone that will be negatively effected by over-consumption and the climate change it causes, such as a truly representative democratic state.
Syndicalism is great for organizing a community and handling it's problems but it can't handle large scale problems that require self sacrifice like war and climate change, for that you need larger organizations.
There are 2 problems with not having enough diversity in training data:
The AI will be worse at depicting diversity when prompted, eg. If the AI hasn't seen enough pictures of black people it may not be able to depict black hair properly as it doesn't "know what it looks like"
The AI will not show as much diversity when not prompted. The AI is working off statistics so if you tell it to depict a person and most of the people it's "seen" are white men it will almost always depict a white man because that's statistically what a person is according to its data.
This method combats the second problem, but not the first. The first can mostly be solved by generally scaling the training data though, which is mostly what these companies have been doing. Even if only 1% of your images are of POC, if you have 1b images 10mil will be of POC which may be enough to train it. The second problem would remain unsolved though since the AI will always go with the statistically safe 99%.
It does, it just favors the dominant ethno-religous complex. Much of the western proffesional dress code has basis in christian ideals of modesty. These cultural signifiers don't occur to us though as they're so normalized. If you came to work dressed like Angela from the office you wouldn't be cited because the dress code was written with that attire in mind and people view it as normal. You'll be cited if you violate those ideas of modesty, eg. Showing midriff, or having different views on modesty, eg. A head scarf.
If you want to say it's completely neutral you'll have to exorcise all christian biases and assumptions from western culture, which they dont seem to be doing here.
Saying this is like saying your a particularly beefed-up bacteria. In both cases they operate on the same basic objective, survive and reproduce for you and the bacteria, guess the next word for llm and auto-complete, but the former is vastly more complex in the way it achieves those goals.
Looks like somewhere close to chabot park near Oakland California. It's overlooking the bay and across to San Francisco. You can see two dogs fucking, then southern Oakland, then the thin strip of water separating Alameda island, then the larger bay, than the southern part of San Francisco.
I could see this going either way for emissions. On one hand people being forced to go into the office requires them to make a daily commute which adds emissions. On the other hand though people being able to work remotely encourages even more sprawl and car dependence for other every day tasks. If you start working remotely in the middle of nowhere you might be commuting less but driving further for other tasks.
Making things biologically and making things with silicon and plastic are two completely different animals. Were far off from making something to simulate the brain in all it's parts, much less making a biological one from scratch. Same thing with a chicken leg, we can make a robot chicken leg that does all the functions of a chicken leg, but you probably wouldn't want to eat it.
Those are mostly for sprinters and short distance runners as they are higher impact and use running cleats with less cushion then softer distance running shoes. They're also more likely to fall in events like hurdles and it's better to land on that then on concrete. Most of the shock that could be absorbed by those tracks could also be absorbed by good running shoes. Most events above a mile take place on the streets.
Sorry to burst your bubble but it's decades away if it's even possible. The current process involves putting animal cells in a bioreactor(vat) with nutrients and having them propagate. It's hitting hard limits on scaling though because the larger the vat the harder it is to get waste out and nutrients in without some sort of vascular system. Even if it did scale it's not producing steaks or even meat chunks, it's just making meat cell slurry that's mixed with a bunch of other stuff to make something like beyond burger but with some actual "cow" cells for probably 3 times the price.
It took evolution billions of years to efficiently make complex multicellular structures, humans are a long way off.
How do they do this without running a foul of regulators? People are already mad at meta and want to break them up for having instagram and Facebook, if they add the last big social media platform every politician right and left will be lining up to take them down. There's a reason they never bought twitter despite being able to 10x over. Combine that with new EU interoperability laws and there's no way meta could get away with that.