The problem is that quite often the money earned abroad isn't actually being taxed at all (or at an extremely low rate). Things like the Double Irish actually prevent anyone from taxing them (or taxing a small portion of the total income), using various legal loopholes, even if the income would normally be taxable. So you can have goods produced in one country, sold in another, and never be taxed anywhere (and in fact the company may well take a deduction on business expenses from production, or other such nonsense). That's the problem here: companies are using loopholes to earn money in countries and not pay taxes on it at all. It's legalized tax evasion.

If and only if the British taxes are less than the taxes you'd be paying in the US, and then only up to the difference. Or you can take a $97,000 dollar exclusion on foreign income (so if you make less than that, you pay nothing). Really, the whole system is only intended to make sure that rich people and corporations that have the money and resources to take advantage of loopholes still pay what they owe (doesn't always work, of course, but thats the intent). Slashdot ought to be all over that.

Quite right. Because no other form of transportation takes place in a vacuum. Unless you know of some radical new physics, standard reaction-mass engines will be necessary for spaceflight for... well, forever, so, I'm not sure exactly what your point is. And yes, they've worked on the idea before with NERVA. We have, believe it or not, made a few technological and engineering breakthroughs since then (mind you: NERVA worked. It worked very well. It was canceled for political reasons, not practical ones).

Ding ding ding ding! We've got a winner here! One Godwin'ed Slashdot thread! (ok, I'm not actually sure you won, but I didn't see anything up above).

The goal of these kinds of security measures isn't to prevent people with malicious intent from breaking them. That is, obviously, impossible. It is to make sure people without malicious intent don't engage in activities which are indistinguishable from malicious activity. That in turn means that if you see people engaging in apparent malicious activity, you can safely assume they are, and operate accordingly (i.e. shoot them, arrest them, etc.)

You might be able to: my Logitech Mx518 (at least with the software I have installed) doesn't seem to allow rebinding the two primary buttons, though, so it's not a sure thing. The OP can do what you suggest, or just ask around (on Reddit, user reviews, even send a message to Logitech).

Aside from the fact that that wouldn't work anyways (intent to link to the illegal material would be required, and that certainly wouldn't meet that qualification), the hyperlinking charges were dropped. Yeah, the Slashdot summary is a bit deceptive (absolutely shocking, I know).

The FLRW metric (which is what the equation that governs the cosmological expansion of spacetime) has a cosmological constant term in it, initially placed there by Einstein to maintain a steady state universe, but which we now know drives an accelerating expansion of the universe. This constant term is exactly that: a constant (negative) energy per volume of space. More space means more total energy.

However, TFS and TFA (I've only scanned the referenced paper, but that looks much more reasonable) are absolutely wrong about why this is a problem. It is a problem, but only in the sense of figuring out where it comes from (i.e. what exact mechanism drives the creation of this energy). The fact that energy is not conserved violates no law of physics: in fact, general relativity doesn't conserve energy anyways, and the expansion of the universe certainly does not (even without the non-conservative nature of gravity).

See, the conservation of energy is a result of Noether's theorem, which states that for any differentiable symmetry of the action of a physical system, there is a corresponding paired conservation law. For time symmetry, this is the conservation of energy. However, time on the scales of the universe is not symmetric. There was a beginning to the universe (which alone breaks the symmetry: you can't shift backwards in time more than ~13 billion years), and the universe as it is now looks nothing like it did 10 billion years ago. So we don't expect energy to be conserved in the universe as a whole (even if it is on local scales).

Dunno what cables you're using, the HDMI cable hooked up to my monitor is only slightly thicker and stiffer than my mouse cable.

Absolutely impossible? No. But the problem is that we don't even know where to begin creating a true AI, which means we also know nothing about what threats it may or may not pose... so we also have no actual way to address those threats. All we have right now is pure, 100% complete speculation (no different from speculating about what would happen if we had FTL travel, or psykers, or met aliens). There are plenty of actual threats to humanity that really exist right now (or could be created with our current knowledge and technology), which makes worrying about something we know literally nothing about kind of silly.

Aye. Now try controlling the engine and fuel injection system, and achieving combustion, without using a spark plug. Because that's what the story is actually about. They're using compression-based ignition (like a diesel engine) rather than spark based ignition (like a conventional gasoline engine), which requires a detailed knowledge of the state of the engine at each cycle.

Nah, it's interesting because it's an application of machine learning algorithms applied to an actual physical problem that might have real-world practical use (the whole Rasberry Pi/Linux angle is a side note in the paper just to show that the algorithm is fast enough for real-world application). A possible 30% fuel efficiency increase in all/some new cars? Car makers would certainly be interested.

That depends. If you should have known the gun would have a significant chance of hitting someone, you could well be facing a full murder charge. Randomly shooting a gun in a field in the country? Probably manslaughter. Doing the same in a crowded shopping mall? Yep, that'd be murder. Likewise this bot was shopping randomly on a darknet that has a lot of illegal stuff for sale, and the creator would (absolutely should have, anyways) have known that, which means he would be legally liable for the purchases (if the government decided to press the issue).

The link the GP gave is to the paper linked directly to by the summary (the direct link to the abstract), so some confusion is understandable. In the future, maybe make submissions discuss one and only one paper (or make it obvious they're two papers)?

From TFA, it sounds like the functions in question (pow() and exp()) work by first looking at a look-up table/polynomial approximation technique to see if the function can use that to get an accurate-enough value, then running through a series of multiplications to calculate the result if the table/approximation method wouldn't be accurate enough. Since this work improved the actual calculation part, my guess is that it will improve quite a few cases. TFA does say the lookup table method works in the "majority of cases", though it doesn't say exactly how big a majority, so it's hard to say exactly.

Right, but we want to continue the "Moore's Law" speedup of processing year over year. And that simply can't happen with single core processing: clock speed is already near the physical limit (as in we would need to start violating the speed of light to increase it much further), and manufacturing process size can't continue shrinking indefinitely either, no matter how close we are to the actual physical limits there. So unless we invent entirely new computing systems (e.g. quantum computers), the only speed gains in the future will inevitably be from parallelization, and there are (for many cases) still massive speed gains to be made in that field, simply because the software was never designed for any parallelization at all. Granted, that'll hit a wall where you can't split tasks up anymore as well, but in many cases this process hasn't even started.

You're quite right about the graphics, though: the long-term future of graphics technology is probably ray-tracing, and that takes absolutely massive amounts of completely parallel CPU power.

That's just the kind of thing Hitler would have said!