You know, technically they
But seriously, if you completed 4 years of a theoretical physics Ph.D. but think your MBA was just as challenging... is there a reason why you were able to complete the MBA is less time (presumably)? It's because PhDs in Physics are harder (and possibly not as well compensated as a similar amount of non-Physics-PhD effort for someone intelligent enough to attempt a Physic PhD). Which isn't to say you chose wrong... PhDs in Physics take FOREVER.
But an MBA hardly gets you a PhD in economics...
In all reality, I agree with Musk, here. Being a physicist (even if just undergrad) gives you a much better leg up on spotting fundamental opportunities for improvement in technology than an MBA does. It really does teach you how to spot fundamental relationships and what really matters in a system, while giving you a broad toolset for general problem-solving.
I doubt you would have had as much of a fruitful time pursuing your MBA if you hadn't been trained extensively in Physics beforehand.
"portends an end to the incredibly fast reduction in storage costs over the last three decades."
Disagree, it's just taking a turn you're not looking at. Solid state has just really started to take off in the mainstream. As the years go on, it will continue to get faster, cheaper, and more reliable. In a couple short years, we've already broken the $1/gig barrier.
After that... Well, it's hard to tell. Many consumers are already running out of things to store on their computers. Heck, I'm in basically the same boat. Even corporations are getting comfortable "big data" setups for reasonable prices. I wonder how much longer until our storage systems get "big enough" for all but the most intense scientific and global data-mining applications...
For a while in the 1990s and 2000s, disk capacity was getting cheaper and denser faster than transistors were. Going to solid-state would mean a slowing of the rate of storage cost reduction (though there was already a slow-down exacerbated partially by that huge Thailand flood), not an increase. Besides, there are some big problems with scaling down the cell size in NAND flash while keeping the same error rate. If a significantly new technology doesn't rescue flash, we could be looking at an end to rapid cost reduction in data storage, or at least it would be slower than Moore's law.
Which isn't to say I'm arguing that spinning disks will out-compete SSD. I expect solid-state to continue to eat away into spinning disk from here on out. Spinning disks have the big disadvantage of basically being up against pretty hard mechanical limits on latency and seek-time while SSD can improve continually in that regard.
I'm going to bet that these aren't stand alone drives that you can buy and use off the shelf, but units that are installed as part of a system that has a helium supply.
No hard drive is sealed. Not a single one you own or have ever seen. If they were then big changes in elevation would make them break due to ruptured seals and deformed geometry.
Thus, these drives probably have a port for helium inlet so the internal atmosphere can be maintaned. (It would not take much. I'd imagine)
This is concept is actually not new. I've seen old hard drives that were used in commercial storage systems that had an inlet for an inert gas (Argon I think) The storage system had a supply of gas to maintain the atmosphere inside the hardrive, presumably to control moisture and prevent corrosion.
No drives I've ever owned have ever been back-filled with helium, either. Or have ever had 6TB a pop.
Of course I know drives aren't usually sealed. But I find the idea of an external helium supply completely untenable. No one would buy it except maybe a few people who care nothing about cost and all about looking high-tech. It would increase maintenance and upfront costs while adding another single point of failure to the whole system. Way too expensive for dubious gain.
No, there are two approaches that seem reasonable:
1), there's a diaphram or piston which moves (passively) to maintain ambient pressure inside the device while maintaining a helium-tight seal.
2) the drives are built mechanically to withstand whatever pressure differentials are necessary. The easiest way to do this (for the least mass) would be to slightly pressurize the drive above 1 atm.
Helium tends to like to leak out of things. One has to wonder if the power consumption and reliability and speed of the drives will worsen after, say, a decade deployed in the field as the helium gradually is replaced by air. I suppose that has the added benefit for the hard drive manufacturer of a pretty firm drop-dead (or at least significantly reduced performance) date.
But the increased complexity of the technical approach, i.e. cramming more platters (and using fancy technical tricks like using helium) versus just increasing platter areal density, portends an end to the incredibly fast reduction in storage costs over the last three decades.
Another option may be to operate the devices in a soft vacuum (back-filled with a little bit of helium, perhaps). That may further reduce drag. However, I believe the heads rely on an air cushion in order to avoid contact with the platters, so there would be a limit to this.
This is fear-mongering. We are restarting production, and the new Advanced Sterling Radioisotope Generators we have developed produce three times the electricity for the same amount of Pu-238.
I could live with the low range if the darn thing could be 'filled' from empty in the same amount of time it takes to fill my diesel (which, incidentally, has more than double the range of an S, and rarely dips below 40 MPG).
If I'm not mistaken, the fastest charging method for a Tesla is using one of the Superchargers (assuming they're available in your area - the nearest one to me is more than 1200 miles away), which still takes at least an hour to get an 80% charge... and that's assuming no lines at the "pump."
An hour waiting is bad enough, but if there's 2 people in front of me... that's 3 hours before I can get back on the road. Fuck that shit, I gots places to be.
You can swap batteries in half the time it takes to fill a car with gasoline. Standard for all Model S. You're welcome.
Linus Torvalds himself is a single point of failure... People who rely on Linux being updated in a timely manner should figure out what the probability of him dying is or suffering a debilitating stroke. Then, calculate if it's worth bribing him not to take part in risky activities, pay for a safer car, etc.
I would be opposed to any sort of calculus in an equation for popular press. But algebra? Yes. Algebra makes some things easier to understand.
And YES, it will require a little more mental effort for most people, but mental effort is a good gauge of how much someone is learning. In fact, put both the equation and a sentence explaining it. But algebra is sufficient for explaining the vast majority of physical concepts in a compact form.
Math equations are a language. A language we are all taught from middle school. We can and should use it, and use it clearly.
You didn't mention software licenses. CAD software, modeling software, computing platforms can cost thousands of dollars per seat per year. Stuff like Solidworks or Pro/E or MATLAB are incredibly expensive, I can't imagine that it's stuff like hardware that costs the most. And companies probably ARE spending too much on software. They'd be far better served by having, say, industry organizations commission high-quality software (perhaps open-source) instead of paying the annual Solidworks or Pro/E tax. Unfortunately, this is a big collective action problem. But that's not to say it can't be done!
As much as it's fun to pick on management, they're probably right: Engineering software licenses are obscene.
Complex? Have you seen an internal combustion engine and all the mechanical workings? The advantage of the gas turbine electric hybrid approach is that your turbine doesn't need to change speed much, if at all. In a car or a helicopter or whathaveyou, you need to throttle the engine, and that makes things yet more complicated. In a helicopter, the mechanical complication makes the time between servicings very short. Electric motors solve those problems. Lithium-air batteries (if developed to full potential, in, say, two decades) could provide plenty of power and energy to get rid of even the gas turbine.
Control is definitely a solvable problem, and if ANYTHING has progressed incredibly rapidly, it's computer and sensor technology. Propulsion is solvable, as well. I don't see what's wrong with propellers controlled by electric motors (which have very fast response and very long life).
There is, in fact, no problem with feeding people in the developed world. It is, in fact, possible for technology to progress and even become widely available without every single problem in the world being solved.
And as far as infrastructure, well, the other replier handled that nicely.
A frog with wings would be a flying frog, not a bird.
BTW, did you bother clicking through to read the article? At least as portrayed, most of the space is taken up by the body of the car. It looks like a flying car, not like a roadable aircraft like the original terrafugia.
We want to create puppets that pull their own strings. - Ann Marion