Yes. AI companies will keep doing things like this if they think they can just pay out later worst case. If they are forced to start destroying the things built with personal info and data they don't own, maybe they'll start to think twice. Until then, they'll just keep doing what they're doing.
Seems appropriate to dig this account out of hibernation now. I was one of the early editors here, from 1998 or so. Worked with Roblimo for years. Very decent guy, did a good job of lending an adult presence to the group that was sorely needed. I'm sad to see him go.
Um... the entire state of software, games and all, is due to how copyright law is structured.. you do know that, right?
There are plenty of problems that can happily use conviently parallel algorithms. The interconnect latency usually only matters if you are simulating a large memory.
How is this a problem? It seems more like desire for the status quo.
a) Plenty of people gab around me gab about Netflix's shows. The difference is that they can gab about the entire season rather than specific episodes. I heard way more talk about, say, Daredevil or Jessica Jones, than I do about Game of Thrones or American Gods.
b) With DVRs and other on-demand services for shows, even weekly shows no longer have the watercooler effect on Monday.
Some people binge watch them all at once, and that's fine, some people watch them when it fits their schedule, and that's fine (I'll watch 2-3 episodes per week). I'd much rather prefer to have the freedom to choose when I need to watch something, than to be told.
Bitcoin transactions need to be "confirmed" by miners to be put into the blockchain and made official. In order to incentivize miners to do this, the Bitcoin protocol supports a transaction fee in the transaction that goes to the miner who confirms it.
Currently, the floor fee most miners will use is 220 satoshis per byte, or 49,720 satoshis for the median transaction size (https://bitcoinfees.21.co/). That's about $0.66.
If the reward of mining bitcoins drops below the cost of the electricity to do so, it would be balanced by an increase in the transaction fee. Theoretically, the reward of mining bitcoins could drop to zero and transaction fees could still make it profitable by confirming transactions.
More info: https://en.bitcoin.it/wiki/Transaction_fees
That is literally what is said.
Read more than the first line.
You obviously were moving forward without thinking!
Back then, I wasn't sure I wanted to have a registered account on a website for commenting... ha!
Took me until 4 digits were around before I said, "Fine, I'll grab an account".
Get off my lawn!
Think for yourself, but have a look here.
Their statistics suck, even if their principles are sound.
I was stating that in a number of cases, you don't need HPC, you need the high memory instead of the interconnect, because basically researchers write programs that just use the interconnect to provide a large memory.
The problem is that in a number of cases a researcher could easily use HTC, but they follow the fashion of HPC, using more specialized resources than necessary.
Don't get me wrong, there are a number of cases where HPC makes sense, but usually what you need is a large amount of memory, or a large amount of processors.
HPC only makes sense where you need both.
If you have the cash to pay off a credit card, then I can't see any good reason to have a credit card instead of a debit card. To pay more fees? How's that a bonus for you?
I have an American Express Blue Cash card; I get ~$400 in cash back every year, and there's no annual fee.
(In addition, the last I looked, the consumer protection on a credit card was better than a debit card, though that may have equalized in recent years.)
If it's so controversial, why not just get the mitochondrial dna from the father?
Too bad it comes with built in regressions: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=658896
The tree of research must from time to time be refreshed with the blood of bean counters. -- Alan Kay
