"beyond geopolitical to financial interests"
Interesting sentence there. It's a sad state of the world, that financial interests are implied to be more important than geopolitical.
729 posts • joined 24 Nov 2007
Interesting sentence there. It's a sad state of the world, that financial interests are implied to be more important than geopolitical.
Journalists at The Register actually bother to read the source, understand it, and write an article, as opposed to "journalists" on most of the web who just ctrl-C/ctrl-V as quickly as possible.
"recent trends showing a growth in equality"
Really? Where could I take a look at these recent trends?
Android's "flag as spam", and similar functions, are still vulnerable to spoofed Caller IDs. Also, the malicious caller can still just call someone else. Flagging spammers at the telecom level is another game entirely.
"I'm surprised the boss didn't have an accident with the shredder."
Give it time.
After 7, I would expect the characters to finally realize that the Force has galactic history stuck in a repeating loop, and *that* is the problem they need to solve, before they run out of planets.
Also, the Commission is not "retroactively imposing taxes". There are treaties, which Luxembourg signed and agreed to, which state that you can't give preferential tax treatment to specific companies. Which means that this isn't a case of somebody making up a new tax and applying it retroactively; this is a case of someone *breaking the law* and getting punished for it. Which sounds absolutely fine to me.
Ask something. If you don't like the answer, ask again. If the algorithm times out, complain about sneering experts.
You don't run a simulation of every single electron. You run a gross approximation that's computationally cheap and good enough for the simulated humans' senses, and you only run the very fine simulations when you detect that some of the sims are performing quantum physics experiments.
If you're trying to argue that these searches are justified even in a free and democratic country, the comparison with China is not really helping.
Just a quick note, I'd be fine with having a single automated lane per direction, implying that everyone goes at lorry speed, if I'm allowed to not look at the road at all. I'd rather spend three hours doing something productive or sleeping, than two hours driving.
They don't need that much output to communicate, true. They do need that much output to blow up their enemies' stars, though.
1) I like the "rate my driving" sticker. Can we get something similar for human drivers?
2) If you claim that you're "95% of the way there, but the remaining 5% is difficult", then you are using the wrong definition of at least one of those words.
"The implication of Article 20 however, is far greater: It explictily forbids most of the EU organisations incl. the EU commission, the EU (fake-)parliarment, most the EU administrations etc. This is the real importance of Article 20."
Look, I get what you mean. But this is a political statement, not a legal one. If Art 20 really made the EU literally illegal, any number of nationalist organizations would have successfully challenged the EU on it. I know that calling something you don't like "inconstitutional" feels really good, but this is not really how law works.
That's not an obstacle. Germany just has to agree to enforce the decisions of the UPC. The enforcement would be done by Germans, thus respecting the Constitution.
You could argue that this violates the spirit of Article 20, but if you did, the same argument would apply to any number of international treaties. I don't think the courts would agree with you.
So, mr. Governor is certain that the robocar is absolutely trustworthy. I wonder what will happen when a robocar inevitably runs someone over - because, statistically, the chance for this is 100% regardless of how good the software is.
Wouldn't it be better to just recognize that a robocar is successful not if it never causes an accident, but rather if it causes *less accidents than humans*?
I think XP was considered a very stable version of Windows - when compared to 95/98/ME.
Finding a name that could actually be pronounced, instead of "ECMAScript", would have helped a lot.
"To derive a description of my credit rating from all the data about me, the program/filter/macro/neural net/AI must have followed a finite number of steps of sequence, selection and iteration."
Yes, and you can log them. However, there is only a single step. That step is a function call that takes as parameters your profile data - plus several thousand numbers that represent the network's weights. Those are the problem. The function's body is relatively simple; it just does some fairly trivial math on all of those parameters, producing a new set of numbers (which may be larger or smaller than the one you started with). This math is done in a single chunk; no divide and conquer here. This is iterated a small number of times; the final output is your credit rating.
The function does not encode the "reasoning" that brought the decision. That "reasoning" is encoded in the network weights, the thousands of parameters. Unfortunately, those parameters are nameless and have no semantics attached, because no human set them; they were set by the network itself during training. That would already be enough to make the process inscrutable.
But it gets worse. Not only you don't know what each of those parameters mean - they don't even *have* an individual meaning. There isn't one or a few weights that encode "prejudice against black men"; there isn't one parameter that is the weight given to your age. Rather, that information is encoded as relationships between weights. You don't know which ones, or which relationship. Which means that if you try to change one of them and run the function again in an attempt to see what your change did, you will find that the output is different for *all* possible inputs, because by changing a single parameter you have changed the relationship it had with *all* of the others.
Basically, yes, you can log everything the network does, and you can track the calculation, but this gives you absolutely no information on *why* it does it.
"A "no deal" situation based on WTO terms might be bad for the UK, but it will also be bad for those in the EU we trade with. It's a situation no sane person wants, unfortunately there are some within the EU who will push their political agenda over what's best for the economy of the EU."
Right. So if the EU offers the basic "four pillars" deal like it does to everyone else, and May refuses because her voters don't want Polish plumbers, picking WTO terms instead, that's perfectly rational, and absolutely not a political agenda.
It's good that people are at least talking seriously about staging. IMHO, until we get serious about staging, going anywhere will remain too costly and difficult to be done usefully.
Won't work. If you somehow destroyed both major parties, they would be replaced by something absolutely identical at the first election. Maybe at the second election, but not later than that. There are reasons the winning parties are what they are; they are neither insane nor aliens. Nothing will change much until those reasons are somehow addressed.
You're not just missing the point. You're missing the point's entire existence. Nobody is inventing new particles, or anything else; dark matter is not a particle. Or maybe it is, we don't know, and that's the actual point. Conceptually, "dark matter" is a question, not an answer. It's short for "the galaxies are slower than the theory predicts, but we haven't figured out how to change the theory to fit that yet... all we know is that the problem behaves a bit like oddly-shaped mass, except we can't see any."
The idea of dark matter as an *answer* comes from sci-fi, junk science, and the occasional anti-science troll who thinks that by misrepresting the scientific method hard enough, maybe he can justify his existance.
Interesting to see that most comments to this article have a single downvote. Makes you wonder if Battistelli reads The Register.
Ah, but the two cars running bumper-to-bumper _are_ bonded, in a way, aren't they? The one in the back will constantly try to match speed with the one in the front. So they don't touch, okay, but they do stick together. That's a kind of bond, and it makes the two cars behave as a single object in some aspects. And you can call that a particle if you want. After all, the bits inside things we agree are particles don't really touch either.
(car crash comments in 3, 2, 1...)
"Their only option is to become not free, to require paid subscription. If there is a financial arrangement with users, then there is a strong link to the user's identity too. Users wanting to post dodgy material are going to think twice about it, or wind up in jail.
That'll put a dent in their business model."
Well, yes. Too bad? If it turns out that there is no way to run Facebook without causing massive widespread infractions of the law, then Facebook dies. There are a gazillion business models that would be very profitable if not for those pesky environmental or safety laws, what are we going to do about that?
The government has the duty to force business models to include externalities. If at that point they die, it means they weren't sustainable business models to begin with. They can even run successfully for decades and be very liked by users before anyone notices, but that doesn't change anything.
Sort of. In this context, I take "autonomous" to mean "unsupervised", i.e. connected, but not to a human.
Rolling out flawless code is not technically impossible, but there are strong diminishing returns. In the time it takes to create a flawless application, your competitors will have released a crashing piece of crap software, that people have bought en masse, because your application was not released yet and even crap software is better than no software.
Sort of. As a business, you can protect yourself against your ISP failing by getting insurance and/or a second ISP.
Protecting yourself against a major bank failing is a whole lot harder. You can get insurance and/or arrange loans from a second bank... and then, when the sith hits the fan, find out that both the insurance company and the second bank were major creditors of the first bank, and its failure will take them down, and *that* will take *you* down. Or maybe you managed to avoid this trap, but half of your customers didn't, and won't pay their debts towards you, and then *that* will take you down. Or maybe your suppliers fell for it, and won't be able to supply you any more, and *that* will take you down. You get the picture.
That situation can and does become exacerbated to the point where it's obvious that if the bank fails it will take down a disproportionate chunk of your entire national economy with it, most of it people who did nothing wrong, nothing that could reasonably be considered to be risky, and never even thought they had anything to do with that bank.
The smart thing would be to simply forbid by law such situations from happening. The smart thing, unfortunately, was not done; in fact, quite the opposite. The next-best thing is to bail the bank out so that the situation doesn't go nuclear, punish the people who knowingly set up that situation, and THEN forbid by law such situations from happening again. Guess which part of that got done, and which ones didn't.
That all depends on how exactly the bit that states that insurers can seek compensation from manufacturers work.
Wait, does this mean that if I just get the battery mod and leave it on, I'll end up with a phone that is reasonably thick and has awesome battery life? 'cause that sounds more interesting than mods.
The thing is, if you have a Facebook account, and tell them you don't, you may be fine. Unless they find out that you actually do have a Facebook account, and told them you didn't. At that point, you are in deep carp for lying to them, which is itself a serious crime. So it's quite risky.
Re changing your passwords, sure, but that will only protect any new content you make after leaving the airport. Forensic software can quickly download everything on your account within minutes of you givine the password.
Personally, if I had to travel to the USA, I would be concerned... I have a Facebook account, but I only made it because a game required it for online backups, so it has literally no content. Same for Twitter, I only made that account because there was one feed I wanted to follow years ago. That has to be mighty suspicious.
Re driving, I would far prefer for the machine not to have the ability to "get creative", and that kind of responses to be well-defined, programmed in advance, predictable and reliable.
"Laughable by today’s touchscreen, full-colour, apps-driven, high-density standards"
Right, but today's smartphones are laughable by yesterday's battery life and toughness standard. I wouldn't mind a smartphone that won't die in less than a day (even after I've had it for a year or so), and won't break if it falls from pocket height.
You make a game where the optimal strategy is to hit the other player, drop an AI in it, the AI finds the optimal strategy. This says more about the game than the AI.
I suspect that creating AI and quantum cryptography is easier than writing demonstrably bug-free code. If anything, because nobody is going to fling megabucks at you when you say "I'm going to write code that doesn't suck!"
Attempting to legally block people from investigating your system is moronic. Anyone who wants to study your system for illegal purposes won't give a carp about whether the study is legal or not, because they are going to break the law *anyway*. All you're doing is hampering the guys who are actually trying to make your system *better*.
In fairness, the very first attempts at building a "traditional" computer were also quite hard to program.
Obviously a planet of finite size cannot support an infinite population, and therefore there is an upper limit for population. But the boundary set by that argument is so high as to make that argument useless.
The actual limit would be more or less determined by how much food we can produce, and *that* is a whole lot harder to estimate because it depends on many things including future technological advances, something that is notoriously hard to predict. It could be that the limit is ~10 billions or lower, in which case we're screwed, but it could be significantly higher. It could very well be that it's high enough that we'll bring everyone at a western standard of living, with its associated stabilization of growth, long before that. Or maybe not. Believe what you will; there is really no way to predict this. But discussions about infinity don't really prove anything interesting.
...like everything else, have their target audience. That is not "everyone".
Automatically defeating CAPTCHAs would be... good, in your world? Are you even aware of the purpose of CAPTCHAs, or do you think web site administrators just like to annoy people?
Well, yeah, but we're not talking about proper AI here. We're talking about what marketing calls AI, but is actually just a set of obscure statistical inference algorithms. People-ethics don't apply to that.
That's a bit like saying that nuclear physics is all about burning cities. Statistics is a fine, rigorous discipline. It's also complex enough that the general populace does not understand it. Which allows some people, politicians and journalists mainly, to make misleading statistical claims all the time without getting called on it. It's also very useful to get people to believe that statistics is an evil discipline, so that they fail to understand it *even more*, and can get bullshitted *even more*.
Freedom of movement of services is supposed to be one of the four cornerstones of the EU free market.
What we would really need is the ability for a third option for each app beyond "allow" or "deny". The third option would be "provide dummy values". Sure, nosy app, you can look at my SMS history, phone status, and emails - only, you'll find I have never received or sent an SMS, my address book is empty, my phone is never used and never rings, and I have no email accounts. I don't think you'll stop working because of this.
My (very wild) guess would be that aminoacids created by non-biological processes can be expected to have an even mix of chirality, while aminoacids synthesized by life forms are likely have a dominant chirality. This should be true even if it turns out that there are bi-chiral aliens, because the two forms would not have the same function and so would be unlikely to have the same concentration.
Users are already inclined to think that computers are magic, right now. Now, imagine adding a few tens of thousands more years of stratified software and hardware layers. I think computer engineers will be renamed Techpriests of the Machine God way before that.
The Caffeine Diet - it WILL make you thinner, and it might not kill you!
"Electric car specialists Kreisel Electric claim the electrified Mercedes G350D comes with a 300-km (185-mile) range and a maximum speed of 183km/hr (114 mi/hr). It can be charged to 80 per cent of its capacity in just 25 minutes and will go from 0 to 100km/hr (62mi/hr) in just 5.6 seconds. In each case, this beats the original gas-powered version of the vehicle."
In each case this beats...? Are they saying that the original gas-powered version of the vehicle had less than 300km range and took over 25 minutes to have its tank filled to 80% capacity?
Wait. This is a case of alternative facts, right?
Very true; "AI" is fundamentally different from traditional software engineering and should never be used to attempt to solve the same problems. You can never rely on the answer of an "AI" system in the same way that you can rely on the answer of a classical algorithm. And this is not something that can be fixed; it's a fundamental property of how such systems work.
That said, the "AI" system should at least not crash outright or allow arbitrary code execution when encountering weird input. Those are traditional bugs and should be fixed as such.
Biting the hand that feeds IT © 1998–2017