You could be right, but it seems to me there is a lot of data going into that thing, if it is designed to not just check the data for matches, but also self-consistency across inputs it could be a pretty high bar... I imagine it does depth, IR temp checking (nose colder than face, etc.), image, and maybe even a couple ms of time monitoring to make sure the face is pliable and maybe even has realistic deformability.
I'm curious as to wether any lawyers here can chime in on how well this TOS would hold up in court. I know when I went to check I was never shown the TOS or even a checkbox that needed to be checked to confirm that I agreed to the ToS with a link to them. It feels like the ToS are what comes with the ID protection service and were meant to apply only to lawsuits that might arise from using the ID protection service, but IANAL.
The first time I learned of adversarial research my initial thought was GD, in our automated AI-driven future, the people who master this are going to live like wizards.
I've been doing preliminary research on what kind of high-density trees grow reasonably fast in different climates. If the wood has a high enough density it will sink in water without treatment. Wood that has been submerged can take 100s of years to decay. Hence it might be possible to lock away a lot of carbon and give us some breathing room, by repeatedly clear-cutting and reseeding forests and sinking the logs. Even without high-density wood, it might be worth the effort to treat the timber so that it will stay sunken. I did some estimates previously that suggest the amount of land area that would have to be dedicated to this to reduce atmospheric CO2 is not unreasonable.
Good. Can't wait to see further editorials in Nature talking about all the problems in science (as a community, career, and funding) whilst the whole time meandering around the elephant in the room: they are a _big_ part of the problem.
Windup girl was an amazingly imaginative extrapolation. I do have a bit of a problem with how fatalistic most post-apocalypse works are. I think what would actually emerge would be very messy, but still far more hopeful than depicted. For example, in a The Walking Dead scenario, the first real settlement they found that was semi-secure would have probably grown into a fairly stable city-state. They work hard on that show to keep the rollercoaster going. IMO, tiring and not very realistic.
I initially thought it didn't look good when looking at the amount absorbed yearly per tree, but then was surprised by some estimates on how many trees you can fit per unit area. A more accurate estimate needs to be done though, different numbers I found put the final range at about an order of magnitude.
Also, most farming and stable forests are mostly carbon neutral. This would require continual seeding, harvesting and sequestration of the cellulose. Maybe it will be a good thing though that cellulose+lignins are so hard to breakdown.
Then again, if my numbers are right, if humankind dedicated <1% of land used worldwide for food to growing trees and locking away the cellulose, we would cancel out worldwide CO2 emissions. Since most of that land is used as pasture, there is more than enough play to keep the world fed, and I could see up to 10% being tasked to this purpose , which seems doable in an emergency.
I've been meaning to try and do an estimate related to this. Just how much CO2 could be pulled out of the air by _extensive_ farming of fast growing trees. We could find ways to use as much of the wood as possible (construction materials, etc.), and just sink the remaining wood in cold water where the carbon would stay locked away for 100+ years.
I believe the DOD already has plans to bring online two new computers in the near future that will retake the crown. At least one of them with Nvidia with a big emphasis on P100 and nvlink.
I'd say US news sources strive to hard to be "fair" where their definition of fair is based on a prior of both sides being on opposite sides, but equally distant, from some "center".
I can't go to CNN (what most on the "right" would consider "left") most days without seeing some headline about Trump unaccompanied by a headline about Hillary, in what I can presume is an attempt to be "balanced". The headlines are crafted to make things look equally revelatory. But when you read the articles you find that one is a serious strike (saying shocking things on a mic), and the other is usually a non-story, ie. leaked emails showing the candidate saying pretty much exactly what their public stance has been all along.
I don't think there are many "major" left-wing sites that are as full of hate and fear-mongering as a large population of their equivalents on the right. You are falling for the fallacy that both sides fall somewhat evenly on opposite sides of a normal distribution of "truth" or "balance". This is just not true. Changing times have pushed the overall country further right, but this process has also distilled the GOP down to a core of angry, white christian (males?). I don't think liberals are exaggerating that much when they say if Obama was white (or Hillary were a man), 20 years ago they might have been the republican candidates.
On a lark, I registered as a republican back in 2008 specifically so I could vote for Obama as a republican. That action, however, got me registered to all sorts of conservative mailing lists that initially boggled my mind with how full they are of pure sh*t. At first it was amusing how crackpot and deranged they could be, but it quickly just became sad and/or scary. Since I am fairly liberal, I am also registered to quite a few dem mailing lists, and the comparison isn't even close. The liberal mailing lists and news sources I might describe as having click-bait headlines, or be a bit too bleeding-heart for my tastes, but the conservative side of the aisle is almost pure fear, mis-truths and hate mongering.
I liked the talk, but when he talks about the GIL keeping Python "fast" I have to laugh. It's so slow already, who cares? He is talking about a factor of 1.3 to 2X to remove the GIL in a language that is already 30-100X slower than C/C++.
Underlying a lot of this is the fact that not everyone can perform like Germany, because, by definition, not everyone can be a net exporter. Germany has benefitted greatly from the Euro to get to its position as a net exporter.
Well said. I find it disappointing how many cancer researchers delight in their arcane knowledge of cell line characteristics and will debate heartily about which is the right model for an experiment, so at the end of the day other cancer researchers can review their work and nod approvingly about their choice of cell lines, etc. These cells are exceedingly unnatural, even by the standards of cancer cells, and their extensive use in research now far exceeds their ability to tell us things that are meaningful about cancer in the body (IMO). I understand science requires reproducibility, and (in principle at least) cell line work is more reproducible... but I'm worried were chasing red herrings. If you are using cell lines to do basic research on a very specific subsystems, that is likely OK, but so much research these days is "translational" and involves observing cell-wide changes and reactions.
I've seen this repeated a lot, but have not had quite the same experience with "permanent" performance degradation. Especially if I eventually expand the pool with another vdev. Not sure about ZFSonLinux, but:
1) Having a ZIL helps with this, and in general.
2) ZFS changes strategy depending on how full it is, it spends more time avoiding further fragmentation rather than grabbing the first empty slot. This hit would go away if you get the free space back up.
3) Finally, there is a way[1] to have ZFS keep all the info it needs in RAM to greatly alleviate the times when it starts hunting harder to prevent more fragmentation. It looks like the RAM requirements are 32GB/1PB... so not too bad IMO.