Everspin previously used the crossed-lines writing technique (shown here http://thefutureofthings.com/upload/image/articles/2006/mram/mram-write.jpg), but has now switched to spin-transfer torque based devices. Several other companies are also working on this, so things to improve rapidly. PR release at (http://www.engadget.com/2012/11/14/everspin-throws-first-st-mram-chips-down/)

Eh, even if he had made up realistic-looking data, there were a lot of other red flags: not saving raw data or samples, no one else making measurements, all other groups unable to reproduce results, etc. In retrospect, it sounds like it only went on that long because he was at a private lab, but I see what you mean.

For any sort of data storage the energy barrier between the two states needs to be large enough that the system doesn't thermodynamically fluctuate between them very often. In practice, this means that the barrier needs to be several times larger than kb*T where kb is the boltzman constant. For computation there's not any hard and fast rule about the energy required, but there's lots of practical ones...

You did say that you believe that education is a business and that the more competition a business receives, the better it is for the customer. I just suggested an alternate option that would be more competitive for low-income families. I'm not implying that you're pro-child labor -- I'm giving a counterexample to show that your blanket statement is silly, and that (assuming education is a business) competition to public education is not always better for the customer.

What I'm saying is that this will do almost nothing to help low and middle-class families.

I checked the numbers for the country where I grew up: http://www.charlottelatin.org/admissions/tuition.asp http://www.charlottecountryday.org/admissions/tuition-financial-aid/index.aspx http://www.providenceday.org/tuition . The medium household income is 50 k$, putting private education entirely out of reach for most families.

Sure, that makes sense for businesses, but since when is primary education a business? A lot of people would agree that a primary education is a human right: http://www.un.org/en/documents/udhr/index.shtml#a26 Unfortunately for many families a more economically competitive option would be to send kids to work at age 14 rather than to school. Allowing them to do that wouldn't improve schools either.

I went to a high school where ~60 percent of the students got free/reduced lunch, which meant that their parents made less than 200% of the poverty level. In my county schools are supported primarily through property tax revenues, and most of these families rent and pay property tax only indirectly. It's a moot point though, considering private school tuition runs around 15-20 k$. Families, especially those who are trying to make ends meet, can't spend ~half their net income on a single child's education.

I had a similar case -- also from NC and did Duke's TIP -> IB/AP -> top tier state school for almost free. I think the author of the article is intentionally confusing the testing results which show how the US does on average with how students who actually end up doing science do.

That's all great, but to some extent many students will be good at science even if they go to terrible schools. Similarly, it's worth trying to give most students a basic understanding of science even if they go into another field.

There goes my slight advantage in academia due to a unique last name (a fairly common one in the Ukraine but spelled in an unusual way in English). I always hoped that it would make up for being trivial to cyber-stalk but oh well.

Oops- too bad there is not a way to fix typos after the fact.

So there's 196 papers retracted since 2001? That's far less than the number of papers published in my subfield (condensed matter physics) each day. It's simply easier to find the tiny fraction that do cheat now that everything is more readily available.

I'm a graduate student in condensed matter experiment, and I'm not at all worried about my future job prospects. Yes, it is very difficult to get one of the ~10 top-ranked tenure track positions that come open every year in a given sub-field, which would require at least one very-intense postdoc (and to some extent a lot of luck). It's somewhat less difficult to get a tenure track position as a second-tier school, and if you're good at teaching there's plenty of opportunities at smaller universities and liberal arts schools. There are however, tons of companies that hire physicists every year, especially ones who specialize in the more applied side of CME (magnetics, semiconductors, devices, etc). That's what I'm interested in, mostly because I would rather not work 60+ hours a week for the next 15 years. Physics is great fun, but it's not the only thing going on in my life. Outright fraud in physics is astoundingly rare. There's Heinrich Schoen and that's about all I can think of in the last few years. It's not a perfect world -- there are some assholes and also well-meaning people who write papers which are flat-out wrong for one reason or another. However, the vast majority of people who work in the field are honestly trying to do good work. I don't think you have anything in particular to worry about.

Being able to grow well-ordered arrays of bismuth selenide and bismuth telluride nanoplates is a great improvement over the original VLS/Van der Waals growth method developed by Cui's group, in which you could grow similar nanoplates but they were randomly distributed across the surface (it's a pain to work with them since you have to track whichever one you want to use down by hand). However, it's not a huge breakthrough in the field and doesn't put us much closer to any of the proposed devices which would actually use topological insulators. Although they don't show any transport data in the paper the quality of the nanoplates may not be that good based on the ARPES data shown -- the fermi level falls well into the conduction band, and not in the gap as would be required by most interesting applications. Also, a more commonly used technique called molecular beam epitaxy (MBE) can also be used to grow continuous films of these materials across whole wafers, and several groups have demonstrated very high quality films this way.
TL;DR: A nice scientific paper, on an exciting topic, but no major breakthrough. Several interesting uses for TIs have been proposed but they are all very far out, everything going on right now is still basic research. (Full disclosure: I'm not affiliated with either group, but I am sitting in the lab measuring some TI-based devices right now).