legend

How so? And FWIW, my understanding from reviews is this newest release is significantly better than last years QLED.

Any particular reason you would still prefer OLED? I was a bit concerned too, but I did some googling. First, it sounds like everyone supports a baseline version of HDR, that being HDR10. HDR10+ is indeed in competition with Dolby Vision, but it sounds like HDR10+ might be better and stands a good chance of winning. I'm basing that off this site and some others, so it's possible I'm being naive. https://www.trustedreviews.com/news/what-is-hdr10-3294683 So while I suppose its possible HDR10+ might not win, based on that info that doesn't mean you can't get HDR content at all even if Dolby wins. If one wants to wait things out, it may be a while, and I was planning on next year being the longest I'd wait regardless. My 08/09 (can't remember which year) Panny Plasma has had a damn fine run though!

For my apartment 65" is actually pushing it on being too big 3k is a lot though. I was thinking I'd probably hold out till the 2019 model but I might cave if this looks like a future proof slam dunk.

Really? No comments at all? Should I have posted this GG?

I was pretty confident that the next TV I bought would be an LG OLED, but I just ran across this promotional material for the Samsung Q9F series: https://www.samsung.com/us/explore/qled-tv/gaming-tv/?cid=smp-mktg-stc-tv-07182018-122317 Highlights (also from search googling): Freesync VRR 20ms input lag No burn in or image retention Only modestly worse black levels than OLED Much brighter than OLED HDR10+ That's a pretty damn compelling counter point to an LG OLED for a gamer. From what I can tell, it only slightly loses on black level. Are there other reasons to still consider an LG OLED though? Do you guys think that LG OLEDs next year would be able to make up the gap and surpass for gaming? Regarding the VRR, I understand that the HDMI standard for VRR may be different than FreeSync, but would it be safe to conclude that whatever it is would be something that could be patched into firmware for a display that supports Freesync?

I couldn't get through it. They both were like this:

How is this country this fucking broken? What the fuck is wrong with this place!? Everyday day brings another piece of insanity. We're governed by absolute morons put in power by absolute morons.

Is he not under contract?

I think it's weird to be either anti or pro gamer.

It's too bad there wasn't an armed teacher at the convenience store, or maybe a really capable 4 year old. Seriously though, I really feel for the kids of that father. First their schoolmates, and then their father

There is a Mexican/Bagel shop in Providence and they make a mean breakfast burrito. Now I want that.

Often times, things I didn't want to know about Trump. More positively, I've developed a new approach for online learning of visual representation and memory encoding that is aimed at being much more sample efficient for use on a robot. Technically, I've been developing this idea over the past few days. Still needs testing. We'll see how it goes.

Because America won't elect insufferable morons?

If someone thinks you can't get a good bagel outside NYC then they're simply insane. NYC Pizza at least has a more clear style to it. So while I might not agree with the same idea there, it's at least not completely insane.

To be honest, it's not a well written paper, so don't beat yourself up too much A lot of work seems more complex than it is, because a lot of computer scientists are awful writers. I don't consider myself "great" at it either, but I do continue to improve. When you first learn about binary systems, you learn that mathematically, there is nothing special about binary, nor even decimal which is what you tend to be used to. Mathematically, they're all equivalent. So the possibility of using other systems is always conceivable. But physically, there isn't really much reason to do that. Computer hardware tends to be easiest to design with an underlying binary system (charge, or no charge). The article you posted seems to be suggesting that for quantum computers, it may be better to do otherwise; that using a different base will be better for that kind of machinery. Seems entirely plausible, but you probably haven't heard much more because quantum computers are progressing very slowly. If we get useful quantum computers in the next five years (meaning, we would actually benefit from using them to do something) I think we'll be doing well. But that should illustrate how early the field still is. Optimization problems tend to be incredibly hard to solve exactly except for very specific classes of algorithms. All of "deep learning" is one big optimization problem and even with all the compute available that people use on it, it's no where remotely near to being able to ensure an optimal answer. Worse still, the answer deep learning gives doesn't even have any approximation bound guarantees! In this case though, the solution they propose is for a fairly specific class of optimization problems and even then, the conventional algorithm for this class still does pretty well. It's not going to take the world by storm

It's a method for optimizing a specific class of objective functions. Roughly, if I give you a set of items, the objective function scores how good that set is, and what we want to do is find the set that maximizes the score it will give. The method also assumes certain properties about how that objective function works (If those properties are not held, it's unclear whether it will work well at all). It's a sampling based approach so there is no guarantee it will actually find the input that maximizes the objective at all. But it has a pretty good chance of giving an input that scores pretty highly and the only reason to use such an algorithm is when computing with certainty the exact optimal answer to such a problem requires far too much compute.

I'm more optimistic than I was from the teaser.

You're not wrong, but I love Queen and will probably be a sucker anyway.

Jupiter is just gaming the welfare system.

My motivation to buy a PS4 is increasing.

This work isn't nearly as general as the Science News article makes it sound. That and most of the contributions are theoretical analyses. That's not to shit on theoretical analysis (it's incredibly important to advancing what we do), but this is not some magic wand by any means.

*If* Valve comes out with a bunch of games (the fabled "3" set of games) showing that VR can be used for amazing games, all will be forgiven.

Might be time to get it.

That could indeed make sense to me, but then the way the article was phrased seemed like they were saying that whatever raise in price it will be from competitors, it will cause the market to shrink in a counterproductive way in which no one wins. But then why would competitors raise the price that much? They should only raise it to the extent that they're still getting a win, because they're not under any external pressure to raise at all, just the US. But I may have just badly misread what the article was trying to say.

I was asking why it wouldn't be the case that other sellers would keep prices low to take the market. I think I may be confused about which actors are being referenced in that quote. They raise a few: "American sellers, American buyers, Chinese sellers, Chinese buyers, and finally other countries selling.