Re: Strontium Dog
So you're suggesting those who took a look at the trailer and went "f### this, forget it" were merely misinformed, and would have enjoyed it tremendously if only they saw it...?
4733 publicly visible posts • joined 4 Mar 2013
"Well done on completely mission the very point he was making."
I must return the compliment - way to miss my point too. I was commenting on power requirements alone, and as long as boffins are using entire cores as "units" instead something more or less equivalent to a logic gate - to which, based on out current limited knowledge a neuron is functionally most similar to, regardless of its inner complexity that keeps it alive - it's ludicrously pointless to even mention consumption side by side. No, we're not using parallel architecture now the way a brain does. We use stuff that does one single thing at a time, working very fast. Which is why it needs so much energy, especially if we go on to build huge clusters of them trying to mimic a brain. If we'd be using much more parallel but relatively SLOW stuff, akin to many-input gates, they would consume very little power even today, even if the resulting device would appear to process massive amounts of data quickly due to its parallel structure and the sheer number of "gates".
TL;DR: neurons are as far as I know NOT ultra-fast oscillating units, which is the thing that makes electronics power-hungry. Any slow-switching electronics simulating whatever it is they actually do would similarly have a LOW power consumption, unlike the myriad of super-fast cores we build our brain simulators out of today. Comparing those abominations to a brain's power consumptions is still not even wrong.
"The way forward in computing is parallelism. There is no other option."
I seriously doubt it. Parallelism is only good for "data flow" processing, which actually approximates humans acceptably - perceptions going in, actions going out, emotions rattling around inside. Given enough runtime, enough state might even accumulate inside for the occasional "I think therefore I am"; but as far as current general-purpose computing goes, it's incomparably better for anything rigorous and precise even now than any "parallelised" (or even our own, "state-of-the-art") brain will ever be. We're being beaten by any pocket calculator for that sort of thing. Yes, parallelism-powered AI is what you'll need for mollycoddling the apparently endlessly ageing first-world population. But it will be useless* as soon as you need a CAD package, or a VR simulation or, you know, serving up a webpage...
* Bear in mind that in this context "parallelism" is typically understood as "a large number of interconnects between processing units, a large number of which being affected by any information diffusing through the system" and NOT "a large number of specialized processing units performing the same well-defined operation on many pieces of data simultaneously" the way we have in GPUs today.
Actually, I have enormous issues with how he highlights the energy and efficiency considerations. The current units we (they) use to simulate neurons have nothing in common with actual neurons, which are more akin to a simple logic gate. A bazillion of "cores" takes megawatts to run because each of them is incomparably more complex (and more active) than a neuron. Do neurons go "ping" billions of times per second? No? Well then. On the other hand, a bazillion of logic gates takes only watts to run - it's called "one single core". It's infinitely less interconnected internally than an equivalent number of neurons would be of course (and it's not wired for parallel processing), which is why that single core is not much use for AI; but to compare efficiency numbers like this is not even wrong. It's just fucking meaningless.
It's kinda both. More or less it's a miniaturised mobile phone with GPS so it can alert / call a list of preconfigured contacts if the wearer presses the panic button, but it also maintains a data connection to a server and reports / can be queried about most of its settings and the travel log, so IoT is more or less justified too.
Problem is, in the modern world advertising and media is not "based on" but instead CONSTITUTES the properties of any product for 99.99% of customers. Some of us happen to be knowledgeable enough to know better, yes. But everybody else hears "extra security for your kids!" on telly and they know this watch makes their kids extra secure (no doubt about that), then they read "your kids peddled to paedos by security watch" in a tabloid and they know their kids are in mortal peril (no doubt about that) whether any of that is actually the case or not.
Which is not to say this stuff is not badly bugged (I'm sure it's plenty insecure) - but rather that most people don't stand much of a chance of making any determination more sophisticated that the above about most things in their lives. Reading up on things takes lots of time (which nobody has, especially regarding every single thing they might actually need to know about every single thing) and it is ultimately severely limited by not having profound expertise and experience in assorted background topics, the way commentards here tend to have a lifetime's worth of about IT.
For instance, I don't believe there is any realistic amount of reading I could ever possibly do to gain a pertinent insight into the intricacies of protein folding - which might potentially be the key piece of information I might lack trying to make an informed decision on some topic related to, say, nutrition or health. We know why end-to-end encryption is fundamentally different than non-end-to-end, what https protects you against and what it does not and so on - the average punter (or mumsnet dweller) does not, never will, and never could. In the absence of a better and more objective source, they have no choice but remain the plaything of corporate ads and sensation-chasing rags...
"The Monte Carlo tree with value and policy networks in the first AlphaGo was trained on human examples. Its learning algorithm turned out to be quite inferior to the AlphaGoZero's more open-ended self-play algorithm."
Actually, the article - while not being factually incorrect - is incredibly misleading by strongly suggesting exactly what you say - which is, however, incorrect. BOTH machines learned by self-play; the significant difference is that AGZero learned ONLY by self-play straight from scratch, while AG was initially trained by human play up to a certain level: "The tree search in AlphaGo evaluated positions and selected moves using deep neural networks. These neural networks were trained by supervised learning from human expert moves, and by reinforcement learning from self-play". Bad, bad hack!
Interesting, considering for me the killer feature (on other channels like SMS or email) is absolutely nobody being able to tell whether I saw a message or not - disabling "message was opened" replies even if requested is the first thing I do in any email client. Plausible deniability, people. YES, I INSIST.
"But when they make the ombudsperson's job permanent then he/she will have total power over the US intelligence agencies"
Naturally! They'll get entrusted with a very serious-looking briefcase with an even more serious-looking Big Red Switch labelled "Snooping ON/OFF". It might even require multiple simultaneously inserted keys! There will be no antenna, of course, no need - it all works by, uh, quantum entanglement...
"sneaky as I am it's engraved reversed ha ha ha hoo! totally secure!"
Bad news: you unwittingly weakened your own security, considering that in the mirror one would presumably be most likely to use to read a plate on the bottom edge of something, your reversed text would actually read naturally de-reversed. See, this is why rolling your own crypto is always a bad idea...
Terribly sorry, I'm with the "feature, NOT bug" crowd here. I feel no obligation to assist said website with any reports about anything, not any more than I refuse to send anyone "crash reports" - it's MY choice, and the answer is no, regardless of what you justify wanting to hear from me with. You clean up your own damn mess and I take responsibility for mine - any attempts to "collaborate" are NOT welcome and WILL be blocked.
Every single time anyone at all talked about potential dangers a company faces when ignoring the Linux kernel GPL, every single time, they all said that said company could potentially be sued by ANY kernel contributor, and they could lose their license to use the kernel without any discernible practical process to ever gain it back, considering nobody can secure explicit permission from all contributors ever again.
So what exactly is the big surprise that someone actually did it?!? Yes, I can see he's not being particularly nice about it (not that I have any idea how does one litigate "nicely" - you either do or you don't) and there are some (financial) aspects of these cases that some people would definitely find arguable, but why is it such a "no-no" that he actually goes doing what everyone always said any contributor could rightfully do?
Sure, making things easier for any accidental GPL violators is a good thing - for all two of them, because nobody else is doing it "accidentally".
More accurate to look out the window though.
Well sure if I would want to know what the weather right outside is, but who would want to know that when my phone can tell me what the weather is like ten kilometres away, out in a field a good while from the far end of the city, where the "local" meteo box actually is!
I seem to be having some difficulty deciding whether there is a genuine crowd of notable size that somehow honestly never encountered several of what must be the literal top ten of most common computer-related hijinks / "you won't believe the stupidity" anecdotes, or else the sarcasm levels this weekend are just too damn high...
Also called the "not my problem, we have a guy for that" syndrome where people seem to actively forget how to even open a window if they reckon there might be someone they can goad into doing it for them. Often observed not only in office settings but also in households with at least one elderly relative.
That's funny, because in other places it's reported that the data collection cannot be disabled permanently, only until you restart the phone. Unless you take the time to uninstall the thing through ADB...
Not to argue with any of the above (it sounds awful) but remote IR did indeed used to use a handful of carrier frequencies like 36/38/40/56 KHz which are at least in theory indeed incompatible on a hardware level with each other. Practically of course the receivers with the most ubiquitous 38 KHz typically received the neighbouring other two more or less just fine, but that's a different story...
In a mathematical sense, sure you could but really really *only just* - the escape velocity is listed as 0.91 km/s and based on the largest diameter given and the 4-hour period, something at the equator is doing almost exactly 0.5 km/s; considering escape velocity is supposed to be square root of two times circular orbit velocity at the same radius, you need only 0.64 km/s to stay in orbit, so... don't cough, or you're floating away...
"Kaspersky is possibly the only one not yet so compromised by the US agencies, by virtue of being neither American nor British"
While pretty much anything is possible at this point, where does that leave a nominally non-US player like BitDefender...?
And yet, if I were to shoot that same question back to you replacing "advanced AI" with "little grey aliens" you would confidently laugh at me and point out there's no way anyone could have made such a cover-up so impeccably perfect as to leave us with so far zero tangible hard evidence of the whole process of their arrival and presence. Incidentally I do agree, but can you see the problem here...?
"fixed the problems that those systems had."
...as long as "fixing" syncing with anything that isn't necessarily a cloud (SyncML) and handling "notes" (or "memos") on a system level instead of some-app-and-its-cloud is defined as "we can't do that any more (at least in Android)". And they certainly thoroughly "fixed" my Symbian S60 phone's week-or-so standby time to "less than two days".
I find it mildly amusing that as much as I like my Galaxy S2, your description applies almost perfectly (maybe missing "alarm clock occasionally hangs phone instead of ringing"). At any rate, the one driving me positively bonkers* is "screen locks right back up after a failed call" (it's PIN locked - with a LONG PIN).
* Previously it used to be "media volume is auto-set to zero after each call", but that's okay - it turned out that Mapfactor's Navigator was doing that. Because reasons.
"being able to write things is where much of the smart in smartphone comes from"
Apparently for 99.99999999% of the user base, "smart" means playing whatever the modern equivalent of Angry Birds is and sexting. Oh, and Facebook - but what you need for that is a selfie camera, not a keyboard (at least not the one you might think of, just one with two emoji buttons).
"Somehow he knew we would all take to using fingers tapping on a screen instinctively."
No. "We" wouldn't. It may be just my personal opinion (and I'm not much into styluses either), but the lack of a proper keyboard is still the very cornerstone of basic smartphone (un)usability for me. And nobody - absolutely nobody - is willing to make another landscape QWERTY these days, even though I'm slowly at the point where I'd gladly sell my soul for one.
While I'm not a fan of getting followed over multiple sites by a common comment system provider, I'm _several_ orders of magnitude more bothered when a site either decides to just rely on Facebook exclusively (thereby effectively denying me access completely) or implements its own (typically way, waaaaay shittier) comment system and expects me to register and log in with them for the once-in-a-blue-moon comment - on every single one of several hundred sites I might occasionally turn up on and happen to have something to say.
Not that I don't approve of the sentiment, but you are aware that unless you plan to visit the US of A (or do something sufficiently naughty to get them to want to extradite you) you should have much more reason to fear your own GCHQ [or insert alternate domestic Big Brother as appropriate] than the NSA, right? Keeping your data this side of the pond will do nothing to protect you from your very own gubmint...
It may have been smart, but where others might see "they pondered the moral issues and implications and decided to cancel" I see instead "they sensed the public backlash in the making and the PR implications and decided not to open that can of worms". Not because they thought it was inappropriate but because they could see enough of the public thought it was inappropriate. One of those things is a working moral compass, the other is a working instinct of self-preservation. Cynical? Who, me...?
Yes, and every time I see that mentioned, the part that I immediately get reminded of is this one:
"In 2016, Burridge and Linden defined the criterion as the time to reach 0 °C (32 °F), carried out experiments and reviewed published work to date.[1] They noted that the large difference originally claimed had not been replicated, and that studies showing a small effect could be influenced by variations in the positioning of thermometers. They say " We conclude, somewhat sadly, that there is no evidence to support meaningful observations of the Mpemba effect"."