I agree. The voice-to-text is remarkably good: definitely at the point that it has become a tool and not just a toy. (I won't say that it never makes mistakes, but it's accurate enough that you can dictate a text message and only have to make a small number of fixes, making it overall faster in many cases.) The Google Now features also work well (asking relatively free-form questions).

However, the 'embarrasement factor' still looms large: I don't want to use the functionality where it might disturb other people (e.g. at work), and I'm even self-conscious using it when walking around in public. (Yes, it remains ironic that we feel weird talking into our phones.) I also avoid using when my wife is in a nearby room, because of the "What did you say? Are you talking to me?" factor. And of course, I usually don't want to broadcast my activities for all to hear. As a result, I'm not conditioned to use the feature, and I forget to use it even in cases where it would make sense (e.g. home alone).

I guess what I'm saying is that the adoption of these technologies might well be more limited by social convention, rather than limitations in the tech itself. I'm not sure if this is an intrinsic aspect of humanity (that on average people don't like talking to technology, despite what sci-fi has long predicted), or whether this is purely generational, and the next batch of users will be completely comfortable speaking commands to their computers/phones/etc. (in which case, the tech will no doubt have to improve; e.g. in order to only respond to the assigned user's voice).

Well, TFA suggests:

The research has potential uses in creating security barriers that permit voice communication to pass through, and in developing types of sound-based microscopes that could find application in research laboratories and medical practice.

The scientific paper further notes:

Such a high concentration of acoustic energy into a small hole of radius enables sensitive detection of acoustic signals with subwavelength resolution ... the present work not only opens the way to the efficient realization of [near-field acoustics] in fluid ultrasonics and underwater acoustics, but also to the analogous realization in solid-state ultrasonics.

More broadly, results obtained for one kind of wave behavior often have implications for other kinds. I.e.: results in controlling acoustic waves sometimes have implications in controlling/sensing light-fields, or radio waves, or even more esoteric things like electron beams or neutron beams (which are also regulated by wave equations).

This work seems to be based on this high-profile paper from 2002:
Ravikanth Pappu, Ben Recht, Jason Taylor, Neil Gershenfeld Physical One-Way Functions Science 2002, 297 (5589), 2026-2030, doi: 10.1126/science.1074376

Abstract: Modern cryptographic practice rests on the use of one-way functions, which are easy to evaluate but difficult to invert. Unfortunately, commonly used one-way functions are either based on unproven conjectures or have known vulnerabilities. We show that instead of relying on number theory, the mesoscopic physics of coherent transport through a disordered medium can be used to allocate and authenticate unique identifiers by physically reducing the medium's microstructure to a fixed-length string of binary digits. These physical one-way functions are inexpensive to fabricate, prohibitively difficult to duplicate, admit no compact mathematical representation, and are intrinsically tamper-resistant. We provide an authentication protocol based on the enormous address space that is a principal characteristic of physical one-way functions.

Basically, they create a slab of epoxy with a bunch of glass micro-spheres randomly distributed within it. When you shine light through it, the multiple refractions/scattering events lead to a complicated path for the various light beams, which interfere to generate a complicated light-speckle pattern on the other side. This multiple-scattering process is of course deterministic, but in practice it is so complicated that it is not feasible to reverse-engineer the internal structure of such a material. (In fact, the method exploits coherent scattering, and because the light-detector can only measure the amplitude (and not the phase) of the scattered light, the problem is formally 'ill-posed': there is no way to invert the coherent scattering data to obtain the material structure. Instead such problems can only be approximately solved with iterative processes; this can be made arbitrarily difficult by increasing the number of scattering entities (glass beads in this case)...) This is analogous to mathematical one-way functions: in principle you can crack them, but it takes an infeasible amount of time.

Ultimately the 'randomness' (uniqueness of a slab) comes from the inital preparation of the slab: you're basically 'freezing in' the random Brownian motion of the micro-particles. Thermal noise is a pretty robust source of randomness.

These slabs are neat in the sense that you can use them to generate multiple pads. A different illumination condition (incident angle, or light pattern) generates a new one-time-pad (see the paper for a discussion of 'how different' the illumination condition needs to be in order to yield a uncorrelated/unique one-time-pad), so one idea is for people to carry a single physical token and use it to generate different OTPs for different communications channels they care about.

These schemes are not without their downsides, of course, but it's a neat idea to use a physical structure (rather than mathematical function) to generate pseudo-random numbers. (Thes slabs don't require a battery to maintain their state; one could image secure ways of generating two identical slabs at fabrication time, and then giving them to the two parties; etc.)

Well actually I think the term "perpetual motion" isn't particularly helpful. You're right that it's easy to imagine systems that undergo a certain periodic motion without end, as long as they are not disturbed (no energy extracted). Some people call that 'perpetual motion'. Other people might reserve the term 'perpetual motion' for discussions of non-ideal systems (i.e. real systems in our universe), which are subject to incidental effects (like friction). For real systems, there are going to be additional channels of interaction that allow energy to move into other systems, and thermodynamics (entropy wants to increase) thus guarantees that these channels will be used, preventing cyclic/periodic motion from being endless. (One can imagine gas atoms randomly moving around until the end of time, normally that random motion wouldn't be called 'perpetual motion'.)

In Newtonian physics, this is true: the orbit will continue forever. (An unremarkable example of perpetual motion. Unfortunately only theoretical since our universe doesn't strictly obey Newtonian physics.) In relativity theory, the orbiting bodies will emit gravitational waves, which means they are slowly radiating away energy. The orbit will decay and the system will eventually end up in a minimum-energy state where there is no orbit/motion. (The ambient random energy of space-time will have been increased: hence entropy increased.)

I believe there are solutions to the equations of general relavitity which include time-oscillatory behavior but do not emit gravitational waves. (Wikipedia says that an ideal spherically-symmetric pulsating mass should not.) But these kinds of ceaseless motion are different than what is being proposed in TFA.

Right. And if that's all that they were talking about, it wouldn't be that impressive. E.g. one can imagine putting an ion in a magnetic trap, giving it a small push, and watching it orbit around through the trap without end. Call it perpetual motion if you like. (In reality it will be radiating electromagnetic energy and slowing down imperceptibly as a function of time.)

The novel thing about these 'time crystals' is that they would exhibit motion and yet be in a minimum energy state. There would be no way to extract energy from their motion. This also means that the motion would be truly perpetual in the sense that no incidental process could cause the motion to stop, since there is no excess energy to be radiated away. This is the novelty of the new states (assuming they actually exist).

Here's my understanding (I'm a physicist though not in a field at all related to the described work):

As usual, the summary and the article somewhat mis-state the interesting part. They talk about 'perpetual motion' but there are lots of examples of things that move seemingly without end: e.g. a planet orbiting a star. However if you think about it a bit more, you'll realize that those kinds of motion can't be used to get "free energy" and actually are not even perpetual. If you try to extract energy from some kind of bound system that exhibits motion, you decrease the energy of the system and alter the motion. So you can extract energy from a planetary orbit (in principle), but then the planet will have less energy, and orbit more slowly (its orbit will decay). As other posters mention, all kinds of natural processes inherently perform this kind of "energy extraction": e.g. random collisision with space-dust, or tidal interactions between planetary bodies, will slowly alter these 'perpetual' orbits. Even if you imagine a highly idealized system (perfectly rigid objects orbiting one another in perfect vacuum), we have reason to believe that such a system will radiate away energy (slowly) by emitting gravitational waves.

What this all amounts to is saying that the system has some 'extra energy' that could be extracted. In physics we would say that the system is not in its ground state, or "minimum energy state". This is the key phrase that the quoted physicists use which the article doesn't properly explain.

The idea is that a system in its ground state will have lost all the energy it can possibly lose. There is no extra energy left. And, conventionally one would assume that a system in the ground state would no longer exhibit any kind of motion: because any motion is extra energy that could be extracted, obviously. So an idealized orbital system has motion, but is not in the minimum energy state. What Wilczek is proposing is that he's discovered kinds of systems which exhibit motion in their ground state. In other words, the system oscillates as a function of time, and yet one cannot extract energy from this oscillation. Cool!

The analogy to crystals is this: as you cool atoms, they vibrate less and less, and eventually they settle into their minimum energy state. This state is usually a crystal, where all the atoms are frozen into perfect rows. This is the minimum energy state. Interestingly, at high-temperature the system was spatially homogeneous (a gas has atoms all over the place), whereas the ground state has spontaneously broken space-translational symmetry: the atoms exist in well-defined positions and don't occupy intervening points. Thus the ground state spontaneously breaks a symmetry (space-translation). Wilczek's proposed states, if they really exist, would upon cooling to their ground state (no excess energy left) settle into an arrangement where they are in motion. Thus along the time axis the system is not constant/homogeneous. The system has spontaneously broken time-translational symmetry. Hence this is like a crystal along the time axis: a 'time crystal'.

I'm not qualified to say whether this is right or wrong. It would be exciting if true. But it doesn't seem to violate any known laws (e.g. you can't use it to violate conservation of energy, so no 'perpetual motion' in the 'free energy' sense), so it seems possible that these states could exist.

In your analogy, you're talking about a very high-level split that can be done cleanly. One person does the creative work of coming up with a game design (storyline, play control, etc.) without worrying about the underlying implementation details. Then another person can certainly do the engineering and coding work to implement that.

But it should be obvious that for some other problems this won't work. For example, it doesn't make sense to try and split the coding into a "creative coder" (who knows nothing about programming) and an "implementation coder" who turns the creative's ideas into actual code. The creative would toss out nonsensical ideas (like "instead of using vectors, why not use genetic algorithms?"), and then the implementer would have to explain why all those ideas are silly... or else they would just have to ignore the creative type and simply code something that makes sense.

In other words, generating good source code requires someone who knows enough about programming that they can see creative solutions. Their intuition is not separate from their programming talent: their intuition is based upon years of training and experience with source code, math, engineering, and so forth. That's where the good ideas come from.

Coming up with good scientific ideas is similar. Analysing scientific data even moreso. It's only once you have a deep, subliminal understanding of the important concepts that you're going to make substantive progress. Whether a deep understanding of math counts as an "important concept" depends on the field, of course... but I would argue that for science generally, the more mathematical know-how you have, the more informed and powerful your ideas will be.

Unless these logs were doctored (unlikely), then Broder lied. However, the one claim of Broder's that Tesla doesn't try to debunk is the loss of charge from overnight cold. Looking at the graphs, somewhere around mile 400, there is a sudden drop in charge from ~45% to ~38%, with a corresponding drop in estimated range from ~80 miles to ~20 miles (the two are not linearly related, presumably because of the intrincacies of the charge/discharge curve being nonlinear). This seems to correspond to what Broder said, that by letting the car sit in the cold, it lost 2/3 of its range.

This is the one negative thing that may have been true in the NYTimes story. Of course, now that Broder has ruined his credibility, even that must be called into question (did he leave it running in a parking spot for a few hours with the heater blasting? ... actually there is a spike in the 'cabin temperature' right at that point...). As someone actually interested in electric cars, that's the kind of question I would like a proper answer to. So, it would have been nice for Tesla to address it (beyond just saying that they have lots of sales in frigid countries).

Predicting the future perfectly is impossible, but coming up with reasonable predictions is do-able. We do it all the time (from the mundane "I'll be there in 10 minutes" to the more general "science"). Throughout this thread I've seen all manner of explanations given for why it's hard. But I think the real explanation is not "it's hard" but rather "it's not worth it".

Consider an analogy. One might say that building a backup application is impossible: files could disappear halfway through a backup, there could be multiple versions of a file, multiple copies of files could exist, network connection could be lost, etc. But none of this makes it impossible, it just means that one has to design solutions that take into account all of these factors (deduplication, versioning, error recovery, etc.). It's a solvable problem, and one that is solved (to varying extents) in existing backup software. We go to the effort because it's important that backup software work properly.

Similarly for a progress bar. If we really wanted to, we could have an API where the OS monitors the computer's performance, and has ready-to-use metrics for "average disk copy speed" and "average download speed" and so on (even with stats on variation by type of task, time of day, temperature of CPU, etc.). Then when a progress-bar widget pops up, it polls this API, and multiplies a pre-computed task list (# of bytes to download, # files to copy, etc.), and comes up with an estimate. Then it monitors the actual progress, and continues to refine the estimate (averaging over appropriate timescales so as to ignore fluctutions). End result is that the time estimate for a progress bar would be quite accurate most of the time (and when it started getting it wrong, it would be an immediate hint that something went wrong). GPS navigation software is actually a good example of this: it appears to multiply miles travelled*speed limit along the route, add in some padding for usual delays (stop signs), and ends up with an estimate that is remarkably accurate. (Of course it can't account for complete unknowns, but that's fine.)

But is it worth all that effort?

My point is: it could be done. The reason it isn't done is because it's usually not worth it. Or at least, programmers don't think it's worth it. I would personally love to have accurate and useful progress bars in all applications. But, crappy progress bars are not the end of the world... and ultimately someone writing an application probably doesn't want to spend a lot of time refining a progress bar that users only see for 10 seconds every couple of weeks. It's not worth their time when there are so many other things that need fixing. And I would rather an application that is robust in terms of saving data but innacurate in estimating save time... rather than an application that knows exactly how long it will take to create a corrupted file on disk.

To pre-empt nitpicks, when I said this:

The predictions from 2006 are predictions for 2012.

I'm well-aware that 2006+5=2011. I'm trying to be as generous as possible in my assessment. If you make a prediction at the very end of 2006, for "5 years in the future", then you have until Jan 31 2011 for that prediction to come true (and the results should be visible in 2012). Thus, their 2007 predictions have until the very last day of 2012 to be realized, if we want to be generous.

Of course even being generous, their predictions are rather awful.

Let's delve into the details a bit. The predictions from 2006 are predictions for 2012. Have they come to pass?

1. Prediction: "We will be able to access healthcare remotely, from just about anywhere in the world" The prediction describes online health records, and telemedicine.
Reality: There have been some efforts, in some countries, to digitize records. Many have failed, some are moving forward. However, to my knowledge, none of them have gained wide acceptance (nor overcome the huge privacy and legal obstacles). The current level of web-integration of our records today is not much different from 2006. As for telemedicine? There have been a few more flashy proof-of-principle demonstrations, but nothing has become routine.

2. Prediction: "Real-time speech translation—once a vision only in science fiction—will become the norm"
Reality: Microsoft recently demonstrated realtime English-to-Chinese translation. However, the very media buzz about that shows that it is far from "the norm". What we have is just tightly-controlled tech demos, not technology integrated into all of our smartphones ("the norm"). It's likely that existing software will get better (text translation has become amazingly good of late)... but it didn't happen within the 5 years they estimated.

3. Prediction: "There will be a 3-D Internet", by which they seemd to have meant three-dimensional navigation/environments (virtual-reality-like).
Reality: Same as 2006, really. We had Second Life, and we still do. We had 3D video-games, and we still do. In fact, this was quite a silly prediction to make in 2006, given how much was already known at that time...

4. Prediction: "Technologies the size of a few atoms will address areas of environmental importance"; this is a vague prediction wherein they reference "Green Chemistry" as if they invented it (they didn't).
Reality: I don't know how to judge this one, since they didn't really make a prediction. There's been more research in the area of green chemistry. Nothing revolutionary has happened in the last 5 years, though.

5. Prediction: "Our mobile phones will start to read our minds", which they clarify as meaning that "mobile devices and networks to (with consent) learn about their users' whereabouts and preferences"
Reality: We can be generous and say that this has come to pass, in the form of smartphones and their associated ecosystem of apps. As a particular example, Google Now (available on Android 4.1 and later) provides contextual information to the user without the user having to explicitly arrange it. For example it warns you that you have to leave now to get to a particular appointment (based on knowledge of your location, the appointment location, and current traffic). If you're at a bus stop, it automatically pulls up the schedule. These kinds of tricks are neat, and will no doubt become more sophisticated with time.

So, my assessment is that their past predictions are right about 20% of the time.

Exactly.

The letter-to-students suggests that 80-hours should be the regular work-week, that works out to:
16/hours a day 5 days per week, or
13/hours a day 6 days per week, or
11/hours a day 7 days per week.

Assuming 7 hours of sleep, three 0.5 hour lunch diversions, 1 hour for commuting, and 0.5 hours/day for bathroom breaks, this leaves the person with about 2.5 hours/day for everything else: running errands, doing laundry, exploring hobbies, relaxing, etc. This is not a fun way to live, and it's also not a sustainable way to live/work: trying to work that hard inevitably results in people being burnt-out, constantly tired, and not very productive. This is especially true in highly-skilled jobs, where the quality of your work comes down to how alert your mind is, and how creative you are... both of which require rest, relaxation, and time spent on diversions.

The 80-hour week is also a lie. That's not how much the professors worked when they were in grad school. No doubt they worked 80-hour weeks on occasion, and those may have even been productive weeks. But there's no way they sustained that kind of work for the entirety of grad school. When I was in grad school we all routinely worked long hours (more than 40 hours/week), and occasionally crazy hours (80 hour/weeks not at all unheard of). But students who tried (e.g. because of pressure from their supervisor) to sustain crazy 70+ hour weeks burned out incredibly quickly.

The letter was trying to encourage the students to work hard and be passionate, which are indeed crucial for grad school. But by setting an arbitrary and frankly ridiculous rule like "80 hours/week" undermines this message.

The Slashdot headline frames this in terms of "Learning HTML", but it's worth noting that the creators of the game don't view it that way. In their FAQ, they say:

Why "almost educative"? The game might have some educative values, because if you play it you learn things about HTML and the bass rules of programming. But the aim of the game is not to be "educative", it's first to be played, to be fun and enjoyed by everyone. You can eventually learn something but it's a plus... not the ultimate goal.

This article has the title "Tenfold increase in scientific research papers retracted for fraud", but at least mentions some actual numbers:

In addition, the long-term trend for misconduct was on the up: in 1976 there were only three retractions for misconduct out of 309,800 papers (0.00097%) whereas there were 83 retractions for misconduct out of 867,700 papers at a recent peak in 2007 (0.0096%).

Percentage-wise, we're talking about a very small number of papers. They quote one of the authors:

"The better the counterfeit, the less likely you are to find it – whatever we show, it is an underestimate," said Arturo Casadevall, professor of microbiology, immunology and medicine at the Albert Einstein College of Medicine in New York and an author on the study.

While this is indeed true... even if the true number of misconduct cases is ten-fold what they measured, it's still a small fraction of the literature. Of course, any number of fraudulent papers is cause for concern (and we should work to remedy the situation); but these results should not cause us to call into question the majority of published science. In fact it points towards the vast majority of papers surviving scrutiny.

To expand upon this...

If someone's primary justification for decrying GPL violations is that its wrong to violate copyright, then it would indeed be hypocritical to support piracy of closed-source software. More generally, if the moral argument is that intellectual creation endows people with some intrinsic 'control' or 'ownership' of their creative works, then this moral argument applies equally to open-source and closed-source creations.

However, that is not the only argument in favor of respecting open-source licenses. In fact it may not be the most prevalent. Many people support open-source software because they fundamentally believe in the particular freedoms that are espoused by open-source licenses: that end-users should be unrestricted; that end-users should in fact be empowered to completely control their hardware, which means having the ability to see and edit all source-code; that sharing should be encouraged. Under the moral axioms of 'sharing is good' and/or 'users should be unrestricted' it is not inconsistent to encourage people to respect open-source licenses while simultaneously not respecting restrictive closed-source (or all-rights-reserved) copyrights/EULAs/etc.

My point here is not to promote any particular viewpoint. Rather, I'm responding to OP's assertion that it is hypocritical to support open-source licenses while simultaneously decrying closed-source licenses (or even going to far as to violate them). It may be hypocritical, or it may be consistent. (There's no lack of hypocrisy in this world, Slashdot included.) Many things look hypocritical only because one is making an assumption of the moral precepts that should be followed (normally, one thinks people are hypocritical because their morals are different from your own).